ELF>@` @8@ܹ ܹ  ))?(o  ) )$$Ptdm m m QtdGNUAYPQ J{C p"  #@012a@AT% $@   RmHJݼ(y a aף{6}tkm@tOjt#H)M ,,qғ75kΑ(B9 S fQ{`ܪ+', +>@+',H+@>h+',p+ +p',+@ +`',+ +P',+ +@',+>0+0',8+` X+ ',`+$ +',+=+',+ ж+&,ض+ +&,+4 +&,(+ =H+&,P+  p+&,x+ +&,+a +&,ȷ+" +&,+ +p&,+<8+`&,@+<`+P&,h+ +@&,+`<+0&,+@7ظ+ &,+7+&,+ (+&,0+0 P+&,X+x+%,+<+%,+ ȹ+%,й+ +%,+ +%, +  @+%,H+  h+%,p+  +%,++%,+;+p%,+`+`%,+ 0+P%,8+ X+@%,`+ +0%,+  + %,+ л+%,ػ+c +%,+ +$,(+ H+$,P+ p+$,x+[ +$,+  +$,ȼ+ +$,+P +$,+ 8+$,@+ `+p$,h+ +h$,+ +`$,+ ؽ+P$,+@+@$,+ (+0$,0+ P+ $,X+ x+$,+;+$,+` Ⱦ+$,о++#,+ +#, + @+#,H+H h+#,p+H +#,+( +#,+@;+#,++#,+ 0+p#,8+ X+`#,`+ +P#,+ +@#,+ +0#,+ + #,+: +#,(+ H+#,P+ p+",x+` +",+ +",+ +",+ +",+ 8+",@+:`+",h+`:+",+ +p",+ +`",+ +X",+ (+P",0+ P+H",X+ x+@",+ +8",+ +0",+ + ",+ +", + @+",H+ h+!,p+ +!,+ +!,+ +!,+ +!,+ 0+!,8+ X+!,`+ +!,+ +p!,+ +`!,+  +P!,+ +@!,(+ H+8!,P+ p+0!,x+ + !,++!,+ +!,+ + ,+@ 8+ ,@+8 `+ ,h+8 + ,+ + ,+ + ,+0 + ,+0 (+ ,0+ P+p ,X+x+` ,+P +P ,+ +@ ,++0 ,+ + , + :@+ ,H+( h+ ,p+@t+,+ +,+a+,+  +,+ 0+,8+ X+,`+ +,+~ +,+x +,+r +p,+r +`,(+z H+P,P+z p+@,x+  +0,+ + ,+T+,+@ +,+u 8+,@+ `+,h+l +,+ +,+ +,+ +,+p (+,0+ P+,X+ x+p,+f +h,+f +`,+ +X,+ +P, +@+@,H+k h+0,p+ + ,+ +,+f +,+ +,+ 0+,8+ X+,`+a +,+ +,+H+,+7+,+, +,(+\ H+,P+ p+p,x+S +`,+ +P,+@ +@,+N +8,+N 8+0,@+9`+ ,h+ +,+ +,+ +,+9+,+`9(+,0+I P+,X+ x+,+ +,+ +,+@+,+ +, +@+p,H+ h+`,p+ +P,+@ +@,+ +0,+; + ,+; 0+,8+6 X+,`++,+ 9+,+ +,+` +,+` +,(+ H+,P+8p+,x+1 +,+1 +,++p,++`,+8+P,@+@ `+@,h+ +0,+ + ,+ +,+ +,+ (+,0+ P+,X+ x+,+Z +,+ +,++,+8+, +t @+,H+@ h+,p+ +,+x +p,+ +`,+ +P,+ 0+@,8+  X+0,`+, + ,++,+S +,++,+ +,(+q H+,P+p+,x+ +,++,+`8+,+ +,+ 8+,@+' `+,h+M +p,+M +`,+ +P,+ +@,+ (+0,0+g P+ ,X+ x+,+f +,+ +,+X +,+G +, + @+,H+h+,p+A +,+; +,+`+,+ +p,+[ 0+`,8+p X+P,`+ +@,+ +0,+ + ,+ +,+R +,(+5 H+,P+5 p+,x+" +,+/ +,+/ +,+ +,+ 8+,@+E `+,h+ +,+) +x,+) +p,+ +`,+@(+P,0+ P+@,X+# x+8,+# +0,+ + ,+ +,+  +, +@+,H+`h+,p++,+ +,++,+z+,+ 0+,8+ X+,`+ +p,+ +`,+ +P,+ +@,+@ +0,(+  H+ ,P+ p+,x+ +,+ +,+< +,+8+,+ 8+,@+ `+,h+ +,+3 +,+ +,+ +p,+ (+`,0+ P+P,X+@qx+@,+ +0,+ + ,+ +,+ +, + @+,H+ h+,p+ +,+p +,+7+,+`g+,+| 0+,8+ gX+,`+u +,+n +p,+ [+`,++P,+ +@,(+ H+0,P+* p+ ,x+ N+,+ +,+  +,+ +,+@D++ ,++++8+@"P+A+p"+G+`d+ +++X+ +n+@ +s(+ @+H++"+7+#+`++0++ + +8+<@+Kh+`+x+++ +x0+.8+@+e`+.x+p:++0Z+n+@++8+@+`+h+@~++}++`}++|++z`+h+++++P+++s+@++ X+p+:+@+.+C(++h+@x+ 1+s+++X+ p+j+ ++++++(++8++x+V+)+p.+P+@ +p.8+P;+(++0++s+ " +(+x+"+;+$++"+. +PHH++X+++`+p1++@+)(+0+@8+p>+<++I++V+P++@++++ ++ +<(+8+@+@+3H+0yX++`+h+Px+~+++0J+|+++p+p++++ f+++L+`]+++ +N+ +$(+pl8+B+@+-H+pX+7+`+h+x+*+++p+@+++@+ ++++` ++++`++++* +(+8+*@+H+X+*`+h+`x+*++0+*++p+*+++*+++ *++0+ * +(+8+|*@+H+0X+q*`+h+0x+@h*++,+ Z*+~+p+M*+v+0+E*+q++ ;*+P+Ц+4* +K(+P8+ "*@+gH+X+ *`+h+qx+*+++@)+>+A+)+ ++)++p+`)+++@)@+ZH+ `+Hh+Px+)+))) )  ) ()0)8)@)H)P)X)"`)h)%p)x))))))5):);)<)>)?))))) )))c)e) )g()h0)8)l@)nH)P)X)u`)zh)p) x))))))))))))))))))))) )()0)8)@)H)P)X)`)h)p)x)))))))))))) ) )) ))) )()0)8)@)H)P)X)`)h)p)x))) )!)#)$)&)')))()))*)+),)-).)/))0)1 )()20)38)4@)6H)7P)8X)9`)h)=p)@x)A)B)C)D)E)F)G)H)I))J)K))L)M)N)O)P)Q)R)S )T()U0)V8)W@)XH)YP)ZX)[`)\h)]p)x)^)_)`)a)b)d)f)i)j)k)l)m)o))p)q))r)s)t) )v()w0)x8)y@){H)|P)}X)~`)h)p)x))))))))))))))))))))) )()0)8)@)H)P)X)`)h)p)x)))))))))) ))))))))))) )()0)8)@)H)P)X)HpH5R)%T)@%R)h%J)h%B)h%:)h%2)h%*)h%")h%)hp%)h`% )h P%)h @%)h 0%)h %)h %)h%)h%)h%)h%)h%)h%)h%)h%)h%)hp%)h`%)hP%)h@%z)h0%r)h %j)h%b)h%Z)h%R)h %J)h!%B)h"%:)h#%2)h$%*)h%%")h&%)h'p%)h(`% )h)P%)h*@%)h+0%)h, %)h-%)h.%)h/%)h0%)h1%)h2%)h3%)h4%)h5%)h6%)h7p%)h8`%)h9P%)h:@%z)h;0%r)h< %j)h=%b)h>%Z)h?%R)h@%J)hA%B)hB%:)hC%2)hD%*)hE%")hF%)hGp%)hH`% )hIP%)hJ@%)hK0%)hL %)hM%)hN%)hO%)hP%)hQ%)hR%)hS%)hT%)hU%)hV%)hWp%)hX`%)hYP%)hZ@%z)h[0%r)h\ %j)h]%b)h^%Z)h_%R)h`%J)ha%B)hb%:)hc%2)hd%*)he%")hf%)hgp%)hh`% )hiP%)hj@%)hk0%)hl %)hm%)hn%)ho%)hp%)hq%)hr%)hs%)ht%)hu%)hv%)hwp%)hx`%)hyP%)hz@%z)h{0%r)h| %j)h}%b)h~%Z)h%R)h%J)h%B)h%:)h%2)h%*)h%")h%)hp%)h`% )hP%)h@%)h0%)h %)h%)h%)h%)h%)h%)h%)h%)h%)h%)h%)hp%)h`%)hP%)h@%z)h0%r)h %j)h%b)h%Z)h%R)h%J)h%B)h%:)h%2)h%*)h%")h%)hp%)h`% )hP%)h@%)h0%)h %)h%)h%)h%)h%)h%)h%)h%)hHp*H /)HHH50HEH(H81S1HPH5^+GHHI+u*H&e+e+Hie+H5]+1HHI+u*H&Qe+77e+H e+CHI+H5J+1HH$I+u*H;&e+rd+Hd+H5[+1eHHH+u*H%d+d+%Hd+H5X[+1HHbH+u*H%od+Ud+0H>d+aH5/[+1HH H+u*H`%&d+ d+;Hc+H5O+1HHG+u*H%c+c+FHc+H5Z+1AHHWG+u*H$c+zc+QHcc+H5tN+1HHF+u*H$Kc+1c+\Hc+=H5;N+1HHF+u*H<$c+b+gHb+H5M+1fHHLF+u*H#b+b+rHb+H5^+1HHE+u*H#pb+ Vb+}H?b+bH5O+1HHE+u*Ha#'b+  b+Ha+H5^+1HHAE+u*H#a+a+Ha+H(1HHH9HHD+u*H"a+/ra+H[a+~H5a+1HHD+u*H}"Ca+_)a+Ha+5H53Y+1HH-D+u*H4"`+`+H`+H5ZP+1^HHC+u*H!`+ `+H`+H5N+1HH{C+u*H!h`+ N`+ʗH7`+ZH5O+1HH"C+u*HY!`+ `+՗H_+H5L+1HHB+u*H!_+ _+H_+H5M+1:HHpB+u*H _+ s_+H\_+H5M+1HHB+u*H~ D_+*_+H_+6H5S+1HHA+u*H5 ^+^+H^+ H5[M+1_HHeA+u*H^+^+ H^+ HHHHHB+u*Hn^+T^+H=^+` H5>S+1HH@+u*H_%^+1 ^+"H]+ H[+H5Q+1HHh@+u*H]+9]+-H]+ H5Q+19HH@+u*H]+<r]+8H[]+~ H5lA+1HڿHH?+u*Hz@]+F&]+CH]+2 H5@D+1HHZ?+u*H1\+\+NH\+ H5K+1[HH?+u*H\+a\+YH}\+ H5H+1HH>+u*He\+fK\+dH4\+W H5H+1HHO>+u*HV\+i\+oH[+ H5+1 HH9+u*H`X+FX+H/X+RH5`>+1HH9+u*HmX+W+ HW+ H5gT+1{HHA9+u*H$W+W+HW+H5D+12HH8+u*HW+kW+HTW+wH5UD+1HH8+u*HR+R+HR+H5H:+1LHH"3+u*HR+ R+HnR+H5F+1HH2+u*HVR+#SHHH9H58+H=(&HHsH58+LRHHeH9u&H5c8+LHL$RHIHL$u2EH5=8+HHD$HL$uE1E1E11I$H58+HHL$+HL$I$H57+HL$H57+L++H=11HH+ HHH+H(8 H%+H5$+H=7+eHF 7+6+11H6+HDHDH=?7+tLH=%7+tH 6+6+H=56+yH= 7+Ht7HH6+u'c HuH(H5H8A1H=6+4HL6+26+11H6+;HHu)HN6+5+1H5+H5HHuEHA6+H5Hy+H5+5+11H5+H=+HH2+u'Hz5+`5+HI5+H=,+HH2+u'Hu;5+!5+H 5+ H=>&+yHH?2+u'H64+4+H4+ H=,+:HH1+u'H4+A4+H4+[ H=,+HH1+u'H~4+d4+HM4+ H=Q,+HHR1+u'Hy?4+:%4+H4+ H=,+}HH1+u'H:4+3+H3+ H=C+>HH0+u'H3+3+H3+_ H=$,+HHu0+u'H3+h3+HQ3+ H=e++HH.0+u'HC3+)3+H3+ H=f,+HH/+u'HM3+2+H2+ H=%+BHH/+u'H2+2+H2+c H=.+HHY/+u'H2+l2+HU2+$ H=++HH /+u'HG2+@-2+H2+ ]11ۅ)H(H=*HH/+H/+H/+H/+H/+HQHH+H1+H7H+H8H+Vy'H1+1+'Hp1+? H=S*uH(H9*uH*H5#1+H=*y'H|B1+(1+.H1+ H5,+H=~1+H*Ry'H70+0+/H0+ H*H=i*H/+H+Ht0+HH+Qy'H0+i0+3Hk0+: H5/0+H=*y'H^0+iD0+7H-0+H=*y'H}'0+i 0+8H/+H*H=*H.+y'H8/+/+:H/+H=t*uHC(H9*uH*H=4*=y'H/+p/+AHY/+(H*H=6*Ho-+H(+H.+H0H+H|H +HޖH+HH+HH+HH+HH+y'H+.+J.+KH.+sH=*uH6(H9*uH*H57.+H=h*y'Hv.+J\.+RHE.+H=)*y'H?.+J%.+SH.+H*H5-+H,+H +H-+HHH +HH=P*H +H*H;*fy'H-+-+ZH-+OH=#*uH(H9*uHz*H5-+H=*y'HR-+8-+aH!-+H=*y'Hq-+-+bH,+Hn*H=pH*+{HHu'H,+ ,+pH,+_HH5'ApHHH++u'H1~,+ d,+yHM,+ H uH+H=HHu'H6,+,+{H,+HH5A HsHHU++u'H++++~H++oH uH菼H=1CHHu'H_++++Hi++'HDH5A HHH*+u'H F++,++H++H uHH=HHu'HG*+*+H*+HH5A`H;HH *+u'H*+*+Hy*+7HbH5UAH HHH)+u'HV*+<*+H%*+HH5A0HHHU)+u'HK*+)+H)+HH5APH?HH(+u'H)+)+H})+;HH5YAHHH(+u'HZ)+N@)+H))+H uHH=&軾HHu'H.)+(+H(+HH5A`HOHH'+u'H(+(+H(+KHH5A@HHH'+u'Hj(+P(+H9(+HH5+1HHtB1HHIu 肽HuH(H57H8׹HMuHɸ 1E15ML%'+u$H'+'+H'+`HH5AHHH&+u$H'+"e'+HN'+H u$H/H115U11LH=ǼHHu'HX'+'+H&+H DH%+H5DHy'H&+&+H&+H Hb%+H5Hoy$H&+t&+H]&+TH H%+H5H*y$HI&+/&+H&+H uHH113H= 蚻HHu)H+%+%+1H%+H "H$+H56Hy)H%+%+1Ht%+H H$+H5HCy)H]%+C%+1H*%+>H H/$+H5Hy)HM%+$+1H$+H DH#+H5Hy)H$+$+1H$+H H#+H5AHey)H$+e$+1HL$+`H H9#+H5Hy)Ho5$+$+š1H$+H fH"+H5Hy)H%#+#+Ú1H#+H H"+H5Hy)H#+#+Ě1Hn#+H HC"+H5HH=y)HW#+=#+Ś1H$#+8H uHH= 趸HHu)HG #+"+ǚ1H"+H H!+H5Hy)H"+"+Ț1H"+H H-!+H5H_y)Hy"+_"+ɚ1HF"+ZH vH +H5Hy)Hi/"+"+ʚ1H!+H `Hy +H5DHy)H!+!+˚1H!+H H' +H5Hy)H!+!+̚1Hh!+|H H+H5H7y)HQ!+7!+͚1H!+2H H{+H5Hy)HA!+ +Κ1H +H H)+H5=Hy)H + +Ϛ1H +H H+H5HYy&Hs +Y +К1H@ +WH H+H5Hy&Hf, + +њ1H+HMu#HٰH1-H1-H= +11DHHu)H++1H+H5v +H=+Hwy)H+h+1HO+sH uH-H=+11CHHu)Hn4++1H+%H5+H=^+Hy)H+++1H+H uH蜯H= +11lCHHu)H+ +Û1Hp+H5 +H=+HUy)H`+ F+ś1H-+QH uH HHu)HR+ +ϛ1H+ H +HHH +HCH1H= +BHHu'H+ +ԛH+H uHkH54 +HL6HHu'Hs+ Y+כHB+fH5 +H=+H'y'Hl2+ +ٛH+%H uH߭HMuHѭ׫HHu)H++1H+H +HHH +HEHH=+JAHHu'H+g+HP+tHMuH-H5 +H5HHu'Ho5++H+(H5x +H=a+Hy'H.++H+HMuH蠬H uH蓬虪HHu)H++1Hm+H+HHH+HCHH=+ @HHu'H}C+)+H+6H uHH5Y+H3HHu'H2++H+H5+H=$+H謲y'H++H+H uHdHMuHVH$HD$HD$ IHH(HtH;-(tHLhL`tHEHPHtHMtIEMtI$H=蝯HImH5HRIHuL貪HuHV(H5H8藫.H(H9Ct,H(H5'H8oH H^1H߱H H+uH;H+HuHH(H5 H8= H+t#H(H5 H81c} Hq+w'H(H5 H81*DuH(H5H81 H(H5H81HIX+e+]HN+H_(H0*H 0+J+H=-5-+([HL$ HT$HLXy$HW++]H+xH5*H=+1HHu$H++]H+7HYH uHuH+z+]Hc+IHLL11:4H<$%H|$%H|$ $1$H +9+H=5+ZHP++H+ H$L$L$H+hHHu)H++1H+H+H5+HHjy)Hu+[+1HB+fHMuHH=x+cH+H *H9Hu)Hx*Ht HH-i*.H=+kHH=+HH*H5I*HHu)H ++$1H+H5+HHHu'H+y+&Hb+HMuH?H+H *H*H9Hu)H}*Ht HHn*.H=+HH=+HM*H5V*HHu)H2++41H+H5 +HHHu'H++6H+H uHeH+H *H-*H9Hu)H*Ht HH-t*.H=+HH=+HS*H5\*HHu)HX++D1H+H5+H'HHu'H++FH+HMuH芤H+H *H*H9Hu)H*Ht HHy*.H=+HH=*HX*H5a*HHu)H}C+)+T1H+4H5T+HLHHu'H=++VH+H uH谣H)+H *H-*H9Hu)H*Ht HH-*.H=6*HH=%*H^*H5g*!HHu)Hi+O+d1H6+ZH5z+HrHHu'Hc)++fH+HMuHբH*H=*1H.*ɡHHu)H ++t1H+H5q+H=+H肩y)H+s+v1HZ+~H uH8(^HHu)HE+++~1H+6H6*H5? +Hy)H<++1H+H#*H5 +H质y)H++1H+H+H5 +Hqy)H|+b+1HI+mH+H5 +H.y)Hs9++1H+*H+H5 +Hy)H0++1H+H*H5` +H訧y)H++1H+H*H5M +Hey)Hp+V+1H=+aH*H5 +H"y)Hg-++1H+H>*H5+Hߦy)H$++1H+H+H5$ +H蜦y)H++1Ht+H+H5 +HYy)Hd+J+1H1+UH*H5+Hy)H[!++1H +H+H5 +Hӥy)H + +1H +H*H5H+H营y)H + +1Hh +H*H5e+HMy)HX +> +1H% +IH9*H5+H y)HO + +1H +H*H5o+HǤy)H  + +1H +H#*H5 +H脤y)H +u +1H\ +H*H5+HAy)HL +2 +1H +=HM*H5V+Hy)HC + +1H +H*H5+H軣y)H + +1H +H*H5+Hxy)H +i +1HP +tHt*H5]+H5y)Hz@ +& +1H +1H!*H5 +Hy)H7 + +1H +H*H5+H询y)H + +1H +HK*H5+Hly)Hw +] +1HD +hH*H5+H)y)Hn4 + +1H +%H*H5+Hy)H+ + +1H +H"*H5k+H裡y)H + +1H{ +H*H5+H`y)Hk +Q +1H8 +\H|*H5+Hy)Hb( + +1H+H*H5R+Hڠy)H++1H+H*H5+H藠y)H++1Ho+Hs*H5\+HTy)H_+E+1H,+PH*H5+Hy)HV++1H+ H*H5+HΟy)H++1H+H*H5s+H苟y)H+|+1Hc+H*H5+HHy)HS+9+1H +DH*H5-+Hy)HJ++1H+HQ*H5:+Hžy)H++1H+H5~*H=+Hy)H+p+1HW+{H uH5H5[H=g1耞HHu)H7++1H+(H8+H5*HHy)HC++1H+H uH蘖H=+ܚH5=*H=+1HHu)H+t+1H[+H=O+HH+Hu.H5*H=`+1yHHu)H0++͜1H+!H=+H+HuЕH5i*H=+1HHu)H(+ +ۜ1H+H=s+Hl+HurH5*H=+1HHu)Ht+#Z+1HA+eH=+H+HuH5*H=F+1_HHu)Hl+$+1H+H=+H+Hu趔Ho*e+@HY*4HU*(HQ*HM*HI*HE*HA*HH:*H=+1H5XH=LHHu)H6+%+1H+'H+H5*HHy)HB+%+!1H+H uH藓H=+ۗH5H=K1֚HHu)H+s+,1HZ+~H^+H5*HH8y)HC+)+.1H+4H uHH=+2Hc*H=*1ݑHHu)H:++71H+H5*H=+H薙y)H++91Hn+H HHH[]A\A]A^A_H= *HH(HtHf.Hh*H=Z*UH)HHw]H(Ht]@H1*H=**UH)HHHH?HHu]H(Ht]H@= *u'H=(UHt H= (h]*fffff.H=ؑ(t&Hߕ(HtUH=‘(H]WKf.HG8HGl~ HGxHÐ1ffff.ATIUHSHHHt HՅu!H1Ht[HL]A\[]A\ff.HHHt H1DHt+Ht&H9 HOHVH9t(\Ht1~\Ht Ð1D1H9HIH#FXL#GXI9uATUSGXH^H9_=H^ H9_ &~mH^(H9_(~ZH^0H9_0~GH^8H9_8~4H^@H9_@~!H^HH9_H~HFPH9GP\St[]A\f.0^`9_`uH_H|HnHtH;HtzHuHtHMH9KuE1HH0HtAH@H9Bu7vtAIcH@HHHH:Hu1H| W1P1H~A1f.HGE1LH@ uLGuLA1LAHHu!10HtH4(HHPHH(1H5o*8HHu910Ht'H*HPHّ(H2HPHHNH HH(1H5*8ffffff.AVIAUIATUSHHHGHH肑L%(p NA; $H oHLLHYx A$OH B~!9'Ht-HH[]A\A]A^R9~@$膑Ht)1H=聐t1H[]A\A]A^HI(H5HD$H:譍HD$fDHGHHt陔fH;=(H;=(t@H;=Q(t鲑fAWAVAUATIUHSHHHID$L$HD$HAHD$H~uIAE1ALHD$8I@HD$(HAHD$ HFHD$0LIILL$8LD$(LHL$ Ht$0LLt$D<$HTLd$Ll$H9uHH[]A\A]A^A_fDH|$HE1H~HL|$Hl$LHLҎIHMHI9uHH[]A\A]A^A_ÐH|$~L9l$uHD$HT$H9uHHHHH[]A\A]A^A_uDAWL~AVAUATUSH(L9HT$L $Ld$`~LHG HIMHD$LHHD$LH|$LIIH $HLL4LHHLHHLٍH\$L9|$~H>(HH([]A\A]A^A_ATIUSHHHoH膏Ht6HPHHtHHH[]A\HfDHH[]A\H(LHD$H:HD$ffff.AULo ATIUHSH^HH931HE11LnHDHTHHHLHI9~H[]A\A]DHGthHGHPHwUH HcHDGHGWHH fG@GWHH H鋍SHHHt1HqH3HVHHtH[HHD$蟈HD$HfAVIAUATIUHSHH#1LLHIPHIt|H1H*H~ LIH4H4HH9u1L草AE I<$HWHI$tAm H[]A\A]A^DLHD$HD$1SHH0HGqH蟐HT$ Ht$HMHH=(H9{HH{HφH{0H,*H*H9tH*H9~yH*H9~dH*H9q~OH*H9~:H*H9~%H*H9b~H~*H9=hH+H<$HT$ Ht$1H{HtHCH/H{ HtHC H/toH{(HtHC(H/tHHCH@H0[DH9{PH/HCPf. ff`fD1fD9։*H @*HcHHwH{HtHCH/tHCH[H@SHtGu蜆t[fDUHHSHφHHtGHEHHH@pPH HQHHtH[]HHD$HD$H[]1fff.SHGHufH藌H{HtHCH/t0HHtHǃH/t#HCH[H@f[fKHtGu蔅f{[HtH/t ff.ATIUHSHHHt HՅuAH{ Ht LՅu/H{(Ht LՅuH{P1Ht[HL]A\fD[]A\ff.H9t;HXHt5HJH~H9rt 1 @H9ttHH9u1DKDSHHGHGHHHHcH1H[GH[DWGHH HcʉH9tH[颩GH[WGHH HHcʉH9uH[+HHcH9nHuHtV@茟HHtHH3HVHH*H߉D$ D$ USHHHt:HHH/tHsH1[]Ht$րHt$H-(HHEH/tHkH1[]D蛀fHGthHGHHwUHHcHD1DGHËGWHH fG@GWHH H˄vfDSHHGtWHGHtDHtHt)x|H[DGWHH H[DH1[GH[HHHtH~H HQHHuHHD$rHD$H$(H5%H8UHffffff.HHIH9H z| H zIH$L H{IHH58LDH5(H81kHfDHxHHxffff.HGHHGHGHHGHGHHGHGHHGAUATUHSHHH9tkHtvHFu X[]A\A]鍯LfM~THvH9t;1H;\t0HL9uE1H\uIM9tJtH9uDH[]A\A]1fff.SHH0HGH识HT$ Ht$H]HHPHHHtBH;(H@@HǃtsHǃHHH<$Ht$HHT$ {HxHtHǃxH/tH}H>H0[|fHHtHǃH/w|HmfHǃ[HG~w0H=1HGH](HHHWtHHQHHt 1fDHH4|1Hffff.AWHAVAUATUSHH$H>Ht$L$HDŽ$HH$~WqHzHHHHD$@I׉4$H|$8|$uGH$HHH$IH$H$H9$uH[]A\A]A^A_HD$H$HDŽ$H@Ht$xHH$~M|$u8HD$xHHH$IFHD$xH$H9$uMWHD$H|$xHDŽ$H@H$HH$~M|$u>H$HHH$IGH$H$H9$uM]HD$H$HDŽ$H@Ht$hHH$~M>f.HD$hHHH$IFHD$hH$H9$|$tHD$H|$hHDŽ$H@ H|$pHH$~M4HD$pHHH$IG HD$pH$H9$|$tHD$Ht$pHD$XH@(Ht$PHHD$`)HD$PHHHD$XIG(HD$PHD$`H9D$Xt|$tHD$Ht$PHD$0H@0Ht$(HHD$H~M3HD$(HHHD$0IE0HD$(HD$HH9D$0|$tHD$1Ld$(H@8HHD$fI$HMe8HH;l$t|$tHD$HX@H~܃|$ tME1Ld$Hl$ L|$@LMLt$8 $HLLIKIm@I9uLd$Hl$ fDL1H HHIU@H9ugMM@MVUHSHHGHHtHHtHH[]HH|(H8ytH/fff.ATIUHSHHHt HՅuQH{ Ht LՅu?H{(Ht LՅu-H{PHt LՅuHx1Ht [HL]A\[]A\ff.UHSHH{(HHH]HHHtH/uvHHHHHHt H/uvH1[]ffff.ATUSHHH9H{{(H9GH9Fu;t7 ~ HFH;C1H[]A\@L%Iz(L9uuL9uҐuH߉vHHH;Az(H;y(H HQHHuH߉D$uD$HKHVH9tHt HVDK DF DD9,A H{0HHA@HDA LN0HHA@LD΃7A 9Ht0LH1x111H[]A\L9HGzfH{HXLNHb7A m7A aHt$rHt$)HHt$rHt$fUHSHHXHh`HXhHpXHP`HHhHt H/u?tHtHmuH+tHtH+tH[]fHH[]tfUHSHH?HhHXH0HPHHHt H/usHtHmuHsHtH+t H[]HH[]sf.UHSHHxw(HHH]HHHtH/u^sHH} HH] HHtH/u4Ic LA<@IUAE\L9PZH53fDI\$(ID$(A}{ID$8HD$L%tPHAD$DID$0ID$8MuE1fDII9LLHIu1H[]A\A]A^A_IAD$Ef.A}:IEtDH8:uLhfDID$I|$(I]HHDpXLQxD1DA)@E7AHwHHuAPЀ SH{A0HJЀ w$f.GHFDBDJA vAlA9~IT$HcHHHTIcH9R),ƒHD<@H`(H5H8^1H[]A\A]A^A_fA|$Dt I|$L$KI|$HL[]A\A]A^A_DLI\$8HAD$DtIL$ 1HHHt HH)IL$ HIE[]A\A]A^A_LID$(ID$ IAD$EID$(ID$0AD$DAD$FfDAH5H_(H81c1[I]LS1MH|$XHD$ID$8IA9uyEAD$GID$(Lk#H1_(H5ʝH8\1[,H_(H5<H81BcH1fIT$H^(H5;HH81HRXcH^(H5DH81bH^(H5H8S\1Hu^(H5H86\1!ffffff.USHHH-_(HHEHkHHHEtH/uZHEH{ HHk HEHtH/uZHEH{(HHk(HEHt H/uZH{PHtHCPH/uZHxHEHxHt H/ueZHHt3H9t#H@@~VHǃtHǃH1[]DHHtHǃH/uY΋0H=cƔ1[ffffff.AWHAVAUATUSHHt$ H6H$L$HDŽ$HH$~bHHIՃ EHD$@t$HrHH4$|$u]EH$t@HHH$IEH$H$H9$uH[]A\A]A^A_H8H/u YHD$ H@HH$~H$HDŽ$H$|$uPEH$t3HHH$IEH$H$H9$u>H8H/uXHD$ H@HH$~H$HDŽ$HD$x|$uJEHD$xt0HHH$IEHD$xH$H9$uTH8H/uXHD$ H@HH$~HD$xHDŽ$HD$pA@EHD$pHHH$IEHD$pH$H9$X|$tHD$ H@ HH$~HD$pHDŽ$HD$h=EHD$hnHHH$IE HD$hH$H9$f|$tHD$ H@(HHD$`~HD$hHD$PHD$X4EHD$XHHHD$PIE(HD$XHD$`H9D$Pu|$tHD$ H@0HHD$H~HD$XHD$0HD$8;EHD$8jHHHD$0IE0HD$8HD$HH9D$0w|$tHD$ H@8HHD$~Ld$8HD$1f.EI$HHD$Me8HD$H9D$t|$tHD$ HX@H~Ճ|$ tmE1LLd$(ML|$@f.L$H$HELIIm@I9uLd$(fDI<$H/s}Ui1ELu6HMfDHM}@H98I?H/u9UH HHIU@H9uH8H/TH8H/TH8H/TH8H/T@SHH HWHHT$Ht$H|$1[HC`HHtlH{H{8VH+H|$HT$Ht$RH{PHtHCPH/t^H{XHtHCXH/t7HCH@H [@ShtClu2H{1RSfDSfDK0HS@E1Hs8H{H @t GHUfiff.AWIAVIAUATUSHH8LgLoHo XWp VP HV(;AD$HHK0HS(HEHHL$ HT$E1ɉD$Hl$E$L1LLWHVx WP HJV(  I9|H8H[]A\A]A^A_@29}V@$fDA|$$CQHtgMuRIcT$HEH9tmHK0HS(AH=QU1D11HK0HS(IcT$L9t&HK0HS(1HuLLPHLLLL7Hffffff.AUATUSHHWH;U(Ht$H;V(HOQE1 Ld$HitzmUHT(p V;P LLHEUp NH B~9|%HtSHH[]A\A]R9}U@$LofHt$aHt$oNUHt51H=IT\1늀tHt$qHS(H5H$H:jQH$NAUIATIUHSHH8HGXHGXHT$ Ht$H$HG`HG`HD$HGhHGhHHD$ FPH{XHt$ Ht H|$WNHD$ HtHH$HtHHD$HHHD$H$IUI$HD$ HEHH8HHT$HhHXHHPHT$ HPt H/uaOHtHmuHMOHtH+t1H8[]A\A]DH(OfD1lH<$IEI$HEH|$H|$ ffffff.ATUSHHH@uotx@to1RHItE1HHUI,$HtuHt)HM@tmHHNHEHPHHUt;[]A\fDH[]A\lN@[HpQ(H5i]A\H8OfD[H]A\ NLNH;Q(H5HH81gUDAWAVAUIATIUSHxsLIHA1H=L*HD 4*D΃HHcHD9D.1f.}8Q9~&)Љ HcHDD9~׉9A9A9~qHcHHD;AuaL!I$H*1LLKHHt hlH"OI,$uLLHt H+<Hx[]A\A]A^A_DLhSHIlLLSIMHɻ*L *E1111l$8Ll$0HT$ HT$HT$HT$H$LL$@1Lt$(LI.IImuLLMNH*AH>D*DǃHcHD9l1@N9)ʉ4 HcHTD9~ωHxH[]A\A]A^A_rKfH=ɺ*HH@XMw`IGXIG`HD$PIGhIGhHD$XQHHH8H׋*H9GH*HH N(H9H;O(tHLMEHt$PIXIW`IGhMw`IwXHt$XHIwhtH7HNHHtYHtH:HOHH t1HtH8HWHHtAAHYJHHD$PJJHD$PHD$XHT$P4JHD$XHT$PH|H=|L1GI}H*AP1E1A9AAE9~IcHHD;jLD;a*DA9}'DHcփHD9HLH\H H\uMcI=*LDhL I$YLZI+1yEfDG1AQLHL(H5۬*1H=r*}M6AX@HHcHJHH**=y* LHx1[]A\A]A^A_QHHE*7*@)*DhL I$oH:L"H/_cHUH5*H=*諻HHHD$`JHt$`H K(HH.HHL$hHT$`HHT$`HL$hH5*HL$`HVLHL$`HH*HH@H*LtNHI1 HK(H 3K(pfffff.ATHUSHL%aK(HI$Lft Gh1tHExHC01tHHC81tHHC@1uPHC(HEHHElC$HE`HCHEXHCEhC HEH{H/t*L9Hkt/1[]A\f.HEpf.FL9HkuHmHCH5Ɩ*H=*1ظHHHHmHvx*^*qHG*1H 9*S*H=ԉ56*1H{H2H/uEHCfDHEJHfDHH`DHtHHD$Hyr1!*K*uH*軾H **H=5*HD$DHHclDHtHHD$H r1*G*tH*KH t**H=Dž5q*lHD$DAUATUSHHH1.?HHHcClLM,M9r[@IM9vNI<$CHHHEH;E HHUHHHEH+uHI@M9wH9?HHMHQHHUuHHD$r@HD$H[]A\A]HclICHHH=f*H@HH3HVHHuHHD$@HD$Hp1*C*tH*HȼHH **H=d5*1MHH?CHpƮ*C*tH*Ho*A*wt1Ho*jHoy*C_*t1HF*AHoP*A6*ut1H*H}o'*C *t1H*ffffff.AUATUSHHH1M9wHHH[]A\A]HHAtHjn*<*$tH*H諺H1衺H ʬ**H=u5Ǭ*Hn*<*"tH*Hm*<*'tHj*Hmw*<]*t1HD*\H5X*H=*1zHHtBH:HmuHH}H9H}PH:HH~H}HyH}XHzHH~H} HyH}`HzHH~H}(HyH}hHzHH~ H}0Hy H}pHz H~aH~(H}8Hy(H}xHz(H~AH~0H}@Hy0HHz0H~HF8HEHHA8HHB8H[H]A\HT_fDL;%1=(=I$I$HuL9[H]A\HHUHHDžHUP~HVHUHQHDžHUXrHVHU HQHDžHU`NHVHU(HQHDžHUh*HV HU0HQ HDžHUpHV(HU8HQ(HDžHUxHV0HU@HQ0HDžHHF8HEHHA8HDžHI|$H^HVhH=| * զ*HH*1Affffff.SHHH}HSlHHH$HH@`ZH$``H9$ RH$+H$h1H$(H9 H$H$pH$0H9H$HH$xH$8H9H$HH$H$@H9H$H~^H$yhH$HH9u[H$~8H$yBH9$Pu8H$~H$yH9$XuH:(HHĠ[ÐHy9(HHĠ[HAf1*tϤ*jwH*胲H *Ƥ*H=z5*1ff.SHHH]HSlHHH$HHH`BHH`H; HBHH`H H9vHOBHH`PH H9?HHBHH`H H9HHBHH`H H9HHBHH`H H9H~nBHH`ysH9 uiH~FBHH`yKH H9u>H~ HcH`yH9 uH7(HHĠ[ÐH6(HHĠ[Hc1i*nO*!wH8*H ,*F*H=x5)*$1ff.USHHGH5*HHHHHEH5*HHHHHLHmt~HCH5S*HHHt~HH}H+t`F6HHH=*HhH:HHH+uH2HH[]H2uH1:HzH^b*h*vHנ*H1蝮H蕮H *ؠ*H=Aw5*tHb*h*vH*Ha*hr*vH[*Hah*hN*vH7*[-9HwHa4*h*v1H*%8Hffffff.UHSHHHGHHKHHHQ0Ht)HHQHHuH0Hz4(HH[]H`*|*oHe*1.H W*q*H=v5T*O1H`J*0*oH*AWIAVILAUATUSHH(Ht$PHD$ HD$0HD$@HHI`E1L$ HAI$IHIcWlHH9sfDH8nHH9wH>3(HHuH6/AM_`EvlH}HUPHuMMDL-HHH([]A\A]A^A_@K7HI(AIą6IGHLLP0H6HHQHHH.@H}HEL\$L}PH$H|$/H4$H|$E1DLD$vD$,L\$H}MH4$DLM-Led/H4$DLLy,H1(H5}*H=J*13HHHH+H*^Ԝ**){H*1lH **H=ps5*H]*r*oH[*+1HHD$ HD$0HD$@HHT$0Ht$ HHLpHH@H $HHH@HL$HL$@xy3HCXHCXHD$ HC`HC`HD$0HChHChHD$@L-ț*D%*L=*+HHT$LH4$mHL$@HT$0HHt$ HD$ HD$0DHD$@-b*D%I*L=2*LH=@r31H,JHt\**%{H*EHM\*ݚ*nnH*H+H\**nnH*HOH[*s*nH\*AVAUATIUHSHH HHD$HFHuoH^HH}HHPHHu*H]1H []A\A]A^H$)H5*LIHV/HHD$IVHCH$H-(H5iL \H 4]HJ]AH811HZ*n*ffHR*H ZH=\J7LnM7Iu#HFHHD$L(HHH\$LBHT$L\H5*LLlWyH5Zߘ*Ř*ffH*R@AWAVAUIATUSHe-HHIEHIEH5*HCHFmIELHHYHH[L= -(L9,HIHEHHhH"(HImI,$H+H:*H Cv*H9HH"v*HHHv*Hk,HI)IEHIEID$H:|*HID$ IM|$(+,HI HXL` LILp(H+uH(HtHmuH'HL[]A\A]A^A_L'H+ H'k&HxXIH],(H0%Y111LL=+(ILM9}_H*Ht*H9PcHt*HuHL5t*Mu +HIIEHIEID$Hz*HID$ HI\$(*HIvLpL` LL.H荌H&<z* z*8Hz*5z*AHT$PL>H5m)LL*9H;z* z*8Hhz*fDAWAVAUIATIUSHHxH-*(HHD$PHD$XHl$`HFHHHV(HC LKI$It$ AH=<(HHD$H q*Huh*Hl$@D$8D$ HL$0H ^*D$H$HL$(H b*HL$Hx*HH;HWHHHHD$HM HD$HHx[]A\A]A^A_H5g*LIHVHHD$PHCHH$HX;H: (H A;H5$IL =HOH81LBH<FH:x*Q x*11Hx*H 9H= RQ 觻1Hx[]A\A]A^A_fDHrLvIt(ItIHF(HD$`HC HD$XHCHD$PLuIIIMMLL$PHD$XHT$`HD$HH<9Hw* w*?1Hw*虅1蒅H w*w*H=Q5w*賺HD$HH5`*LHVV HHD$XtDIMcH5`*LHV, HtgHD$`I4MLH=1;͈Hf8,w*Q w*0Hv*5w*AHT$PL:H5-)LLj5H8v*Q v*1Hv*fDAVAUATIUSHH`L [*H-h (HLL$PHl$XHFHoHHTHH$IHk (H{8H k8H5NFL :HII?H:H8I1AoHH7v* u*00Hu*H !7H=ZO и1H`[]A\A]A^ÐHV LNI$HZ*It$ AH= (HHD$(H l*HD$H^*Hl$@D$8HL$0HL$D$ HD$H$s*HH;HWHH]HHD$HHD$HH`[]A\A]A^@HHH]*HHVb HHD$XIMLL$PHT$XHD$HH5Hot*L Ut*0H>t* 1H +t*Et*H=M5(t*#HD$HKHF HD$XHFHD$P.HrH5"]*HHV H(HD$PIHT$PL7H5)LH92"H4s* s*00Hrs*ffff.AVAUATIUSHHpH X*LpX*H-!(HHL$PLD$XHl$`HFHHHHV(LF LNI$It$ H=a(HHL$(HHi*LD$AHl$@D$8D$ HD$0HB[*D$HD$H.a*H$p*HH;HWHHHp[]A\A]A^HIHTHH$IH`(Hp4H `4H5CBL 6HII?H6H8I1Ad H=3r* q*00Hq*H 3H=K ŴHp1[]A\A]A^fDHHD$HHD$HHp[]A\A]A^ÐHHwLvIt(UItINHF(HD$`HF HD$XHFHD$PHeIIIMM LL$PLD$XHT$`H &V*HD$HH%2Hp* p*40Hp*~1{~H p*p*H=WJ5p*蜳HD$H"H~H5&_*HHV:HtHD$PIMWH5X*HHVHtHD$XIM-H5X*HHVHtHD$`IMLHT$PL3H5)LHl.H1o* o*//Ho*AVAUATIUSHHpH T*LT*H-a(HHL$PLD$XHl$`HFHHHHV(LF LNI$It$ H=Y(HHL$(HH6f*LD$AHl$@D$8D$ HD$0HX*D$HD$H^*H$L 2HII?H]2H8I1AH}/Cn*L )n*//H n*H V/H=GL Hp1[]A\A]A^fDHHD$HHD$HHp[]A\A]A^ÐHHwLvIt(UItINHF(HD$`HF HD$XHFHD$PHIIIMM LL$PLD$XHT$`H fR*HD$HHe.H(m* m*/Hl*z1zH l*l*H=F5l*ܯHD$H"H~H5\*HHVzHtHD$PIMWH5U*HHVPHtHD$XIM-H5U*HHV&HtHD$`IMLHT$PL>0H5)LH*HU-l*L l*k/k/Hk*AVAUATIUSHHpH Q*LP*H-(HHL$PLD$XHl$`HFHHHHV(LF LNI$It$ H=(HHL$(HHvb*LD$AHl$@D$8D$ HD$0HRT*D$HD$HVZ*H$|i*HH;HWHHHp[]A\A]A^HIHTHH$IH'H,H ,H5:L +/HII?H.H8I1AH+j*ij*..HMj*H +H=_DEHp1[]A\A]A^fDHHD$HHD$HHp[]A\A]A^ÐHHwLvIt(UItINHF(HD$`HF HD$XHFHD$PHIIIMM LL$PLD$XHT$`H N*HD$HH*Hhi*G Ni*/H7i*w1vH $i*>i*H=7C5!i*HD$H"H~H5VX*HHVHtHD$PIMWH5 R*HHVHtHD$XIM-H5BQ*HHVfHtHD$`IMLHT$PL,H5/)LH&H)[h*Ah*..H%h*AVAUATIUSHHpH GM*L0M*H-'HHL$PLD$XHl$`HFHHHHV(LF LNI$It$ H= 'HHL$(HH^*LD$AHl$@D$8D$ HD$0HP*D$HD$HV*H$e*HH;HWHHHp[]A\A]A^HIHTHH$IH 'H0)H )H57L k+HII?H*H8I1A$H'f*f*a.a.Hf*H 'H=@腩Hp1[]A\A]A^fDHHD$HCHD$HHp[]A\A]A^ÐHHwLvIt(UItINHF(HD$`HF HD$XHFHD$PH%IIIMM LL$PLD$XHT$`H J*HD$HH&He*e*.Hwe*Bs1;sH de*~e*H=?5ae*\HD$H"H~H5T*HHVHtHD$PIMWH5LN*HHVHtHD$XIM-H5M*HHVHtHD$`IMLHT$PL(H5O)LH,#H%d*d*M.M.Hed*AWAVAUIATIUSHHhH-*'HHD$PHl$XHFHHHV LKI$HX* f1fH +X*EX*H=&35(X*#HD$HH5G*LHVHHD$XtDIMcH5x@*LHVHtgHD$`I4MLH==iHW*W*<,HkW*5uW*AHT$PLH5})LLHIW*/W*F,HW*fDAWAVAUIATIUSHHhH-'HHD$PHl$XHFHHHV LKI$H;*It$ AH='HHD$(H M*HD$HFJ*Hl$@D$8HL$0HL$D$ HD$H$T*HH;HWHHHHD$HHD$HHh[]A\A]A^A_H5I*LIHVHHD$PHC@HH$L /HbHH H5%LOH'HNH8LBH1HU*U*++HeU*H H=0]1Hh[]A\A]A^A_@HqLvIIML,LDMIIu*M~.H5=*LHVHHD$XIMLL$PHT$XHD$HHHT*T*+HT*Qb1JbH sT*T*H=/5pT*kHD$HHF HD$XHCHD$P.HT$PL|H5')LLUH]#T* T*++HS*AVAUATIUSHH`H-'HHl$PyHFHdHHVI$L 8*E1It$ H=@'HHHl$@HJ*D$8LL$(D$ LL$HD$0HD$D$H$Q*H)H HQHHHHD$HHD$HH`[]A\A]A^LfH'H )HH$L kH5"HHHHHLBH LIH'H81 HR*dR*Q+Q+HvR*H H=-dn1H`[]A\A]A^HLnMI-HFHHD$PXHHT$PgHD$HHIH R*Q*w+HQ*_1_H Q*Q*H=#-5Q*HD$HJHHI~H5/:*HHVSHtHD$PIFXHT$PLH5M)LH;HYQ*d?Q*C+C+H#Q*fffff.AWAVAUIATIUSHHxH-'HHD$PHD$XHl$`HFHHHV(HC LKI$It$ AH='HHD$H G*HED*Hl$@D$8D$ HL$0H 5*D$H$HL$(H <=*HL$HN*HH;HWHHHHD$HHD$HHx[]A\A]A^A_H5C*LIHVHHD$PHCHH$HH'H H5L <HOH81LBHHO*{O***H_O*H H=*W1Hx[]A\A]A^A_fDHrLvIt(ItIHF(HD$`HC HD$XHCHD$PL%IIIMMLL$PHD$XHT$`HD$HHHN*_N*+H~N*I\1B\H kN*N*H=)5hN*cHD$HH5B;*LHVHHD$XtDIMcH56*LHVHtgHD$`I4MLH=J}_HM*M**HM*5M*AHT$PLH5])LL HM*oM**HXM*fDAWAVAUIATIUSHHhH-'HHD$PHl$XHFHHHV LKI$H,2*It$ AH='HHD$(H C*HD$H@*Hl$@D$8HL$0HL$D$ HD$H$ K*HH;HWHHHHD$HQHD$HHh[]A\A]A^A_H5@*LIHVHHD$PHC@HH$L oHHBH DH5LOH'HNH8LBHG1F*)H'F*H= WHSF*E*)HE*tAWAVAUIATIUSHHxH-'HHD$PHD$XHl$`HFHHHV(HC LKI$It$ AH='HHD$H <*H8*Hl$@D$8D$ HL$0H l**D$H$HL$(H 8*HL$HC*HH;HWHHHHD$HHD$HHx[]A\A]A^A_H5r8*LIHVHHD$PHCHH$HH'H H5L HOH81LBHHeD**KD*A)A)H/D*H xH=Y *'1Hx[]A\A]A^A_fDHrLvIt(ItIHF(HD$`HC HD$XHCHD$PLIIIMMLL$PHD$XHT$`HD$HHHC*}eC*g)HNC*Q1QH ;C*UC*H=f58C*3HD$HH56*LHVHHD$XtDIMcH5+*LHVHtgHD$`I4MLH=:MTHB**B*%)H{B*5B*AHT$PLH5)LLHYB**?B*/)H(B*fDAWAVAUIATIUSHHxH 2'*H-'HHD$PHL$XHl$`HFHHHuHH$L H?HH H5LOH'HNH8LDH 1HwA*]A*((HAA*H H=91Hx[]A\A]A^A_HV(HK LKI$It$ AH='HHL$HH8*Hl$@D$8D$ D$HD$0H%*HD$(H)*HD$H~)*H$ ?*HH;HWHHIHHD$HSHD$HHx[]A\A]A^A_H8H,LvIt($ItIHF(HD$`HC HD$XHCHD$PL5IIIMMLL$PHL$XHT$`HD$HHH?*%?*(H?*YM1RMH {?*?*H=5x?*sHD$H2M~H5(*LHVHtHD$XIM^H5'*LHVHtMHD$`I/ML6H5'*LIHVHHD$PuHCHT$PLVH5)LL?H>*>*((Hx>*2AVAUATIUSHHpH #*L#*H-1'HHL$PLD$XHl$`HFHHHHV(LF LNI$It$ H= 'HHL$(H 5*H '*LD$Hl$@AD$8HL$HHD$0D$ D$H$<*HH;HWHHHp[]A\A]A^f.HIHTHH$IHp'HH pH5S L HII?HH8I1AtHM=*<*2'2'H<*H &H=WHp1[]A\A]A^fDHHD$HHD$HHp[]A\A]A^ÐHHwLvIt(UItINHF(HD$`HF HD$XHFHD$PHuIIIMM LL$PLD$XHT$`H 6!*HD$HH5H;*z;*X'H;*I1IH ;*;*H=/5;*~HD$HH~H5**HHVJHtHD$PIMWH5$*HHV HtHD$XIM-H5#*HHVHtHD$`IMLHT$PLfH5)LH|H%:*:*''H:*AVAUATIUSHH`L *H-x'HLL$PHl$XHFHoHHTHH$IH{'HH {H5^ L HII?HH8I1AHX:*n:*``H9*H 1H=n|1H`[]A\A]A^ÐHV LNI$H*It$ AH=7'HHD$(H 0*HD$H"*Hl$@D$8HL$0HL$D$ HD$H$7*HH;HWHH]HHD$HHD$HH`[]A\A]A^@HHHiH3'*7HT$LxH5 )LLHm'*Y&*hH&*7H=0v8H+&*Y&*hH&*ffffff.UHSHHGH5*HHHHHEtEHCHHHHt5Ht5H HQHHuHHD$HD$H[]HHHd&*%*dH%*H313H %*%*H=Z5%*h1H%*%*dH%*肾Hf.AWHAVAUIATIUSH8Y10HHH-'H$*HCPHCHEHkHk Hk(HD$HMHD$HEHD$ pID$HHHH$HH'H H5L 5HOH81LBHH$*Yt$*hhHX$*H H=Z YPgH+F1|ID$(HD$ AoD$I|$ Ml$fD$HG!HGHHH)HcH@E1DH|$ HH;=ø'H;=9'DIEH{HHPHH>H!*H9CLkDHsHDLsEH{P *PHS0HCA)HSp1:Ou 1zHS8HCiH*7HT$L(H5)LLXH*Y*hH*7H=&0H*Yk*hHT*ffffff.AVAUIATUS+HHH*H=*HHEH*HL5h *裶HHHIHAHHL/I,$HuL蝮HnHmHCL%^*HHLH5HHHEHH+{IEH5*LHHDHH HCH5*HHHIMH+9H*'I9D$L5x*LLͶHHfHH/I,$HIHXIELH51*HHHHXHCH5X*HHH,IMH+HHHEHhL` Lh(HmuHH[]A\A]A^HЬBHxH谬^H蠬L萬H耬^H**WHi*L1)H1')L)H H*b*H=s5E*@^%&IHd***HE1E1H*HH}'H8襭tL*H**9H*Ho(H1e(1^(H **H=5*]sHy*_*ZHH*>HHE*+*61H*fHV**11H*=H-**JE1H*Q1^HIt!HH'H8贪ImuLuH*z*ME1H`*֮HuIFLuH'H8RH5*L1H\IŐHT**UE1H*xܱHhH**PE1H*Af.AUIATIUSHHHGH;'lH;]'EHHI$L`IELh HCHLMY*L%'p VA;$P 1HHAIA$H H B~E9|KMHEHPHHUHHPHHttHL[]A\A]fDR9}葬@$HWBፃ1 H4$Ll$LBuH{HAI@H(HL[]A\A]f.HPH4$HT$HzTI>H=vE11HH IH=QH1[]A\A]@R9}衣@$H耟 Ht]H*y*eeH]*H=Qܢ1HLuHHH'H59H8ffffff.AUATUSHH/ *L-)HCHH"ϢL%@'x WA;$P 1HLH裢A$H H B~k9|qHHOH+thH *e *HI *H H=APH1[]A\A]@R9}@$HyHt]HY * *H *H=L1HLHHH'H5H8qffffff.AUATUSHH *L-)HCHH"?L%'x WA;$P 1HLHA$H H B~k9|qHHNH+thHE * *eeH *H H=;NH1[]A\A]@R9}聠@$H`Ht]Hs *Y *eeH= *H=1輟1HLUHHHx'H5H8ffffff.AVAUATIUSHHGH5*HHHH HEH5])HHHFHH!HmaHCH5&)HHHHHDH+;H=*L$$HWH;='/H;'7HOQ7E1 HYL-^'H QA;UP bLLHÞAUH H B~s9|yHHkHIH=H*HX HHh訢HH2I,$uL=HH[]A\A]A^DR9})@$yLw%HHHcFH=LH-tH H謡HVH *d*vH*HJH1@L8H a*{*H= 5^*YK?H2HyH9*d*v1H*Hf*d*wvE1H*URH1H,*e*vE1H*H*d*rvE11Hv*H=jHj*dP*tvE1H6*H'H5CH8 OfDAVAUIATUHHcSؚHIH=m'HI講HHHEHhH-*Lp(L` HELM6菛L%'H QA;$P 1HHAHbp VP A$=H~;9|AHtZH+uH$HELHuH [H]A\A]A^ @9}@$L5'IcHH?**xE1E1H*L1xLpHhH **H=d5*HXH=m1HH葝HHfHO*5*xH*gH~(**x1E1H*;H^'H5H8ǖfAWAVAUATUSHHH]L-'HALnH5)LgPIEL9L5M'M9t$L9vA|$ !~ XID$H;FLePH58)L9M9t$L9vbZA|$ /~ fID$H;FxfDM9uuL9uuLAHHH;'AH;3'D/EH*uHuE[E0DHEHHE HCE0C$HE8HC0HE@HC@HC8HEHC HC1Au1HC(HEH{H/t.L9Hkt31H[]A\A]A^A_@HE(f.˓L9HkuHmHCDM9uL9u @L HIH;'AH;'D*EI.uLAEERDIL$HVH9tHt HfEL$ DF DD@9A zI|$0IHA@IDA VLN0HHA@LD΃7A 9PHLE1HgA@L9HHT$袗HT$ADH(lIL$HVH9tHt H%fEL$ DF DD@9A I|$0IHA@IDA LN0HHA@LD΃7A 9HJLE1HgA$LNHI|$HH'H5H836M9LxA7A y7A mLNHMI|$H)7A N7A BH5V)H=/)1HHQHAHm/H))bH)1P H y))H=t5v)qBH{HNH/tmHC8LHt$>Ht$H0))bH)sHHt$Ht$ɏLHt$ڍHt$H"))bH)HHt$虍Ht$HbfHпz)`)bHI)@AWIAVAUIATUSH(HFH5)'LgHH9IIE1E1E11[HoHIEXHHHH H9HIHxI<$IIMHMrI9pM;pI@JIHHt+Hmu$HHt$LD$L $IHt$LD$L $H'H9CHSHBHH HkAElIM`IIExHJ IJHHJ<IUxJ,LL$QHHL$H=)HL$ΕHIL$HmH=<)LL$;HHL$?I,$HL$Y>HmL$H9))NE11H)HL$ L H ))H=[5)>L$H1))ekH)LHE1l 1e H ))H=!5)>HHQHt$LD$L $АHL $LD$Ht$HHt$LLD$L $AHHL $LD$Ht$L$vHL$t%H'H2H9/L$1L$DI(uL誋HtHmuH薋H(L[]A\A]A^A_HH}kCHH \HHt$LD$L $HHL $LD$Ht$HHt$LD$LL$H$6H$HLL$LD$H2HNHt$HH HHt$LD$L $׊Ht$LD$L $gLL$hHHL$H=1)HL$HIL$yHmrH=S)LL$8HHL$I,$/HL$p;HmL$HP)6)}E11H)1M;pHL$ϹL$H))RkHHn)Hθx)^)xHG)?H*L$LL$L$H+))uH)HZ))sE1H)H;t'LH4$rHIH4$tlH@LMt3IH))dkHh)Hȷr)X)7k1H?)HI)/)5k1H)gHv ))IH)HL$ΆL$bH:))FH)H))DE1H)@HIxH~<1L)1HHHHHyPH'H5H8+1Z11^fffff.AWAVAUATUHcSH(L-'H\$`Ht$HT$L9L%q'I$貉HILhH)1IEH=)LHMf(IF HI@I.IHI$Ht$`HtHC@uHHCH5)HHHdHI$xH/I$xHIEI$HHHLSXHs`DKhLCpH{xI$HD$hI$I$8I$MT$XI8I|$xIt$`EL$hMD$pID$HAl$lMl$PI$IDŽ$I\$xH9A$I$I$-I$8HxH8HH9wL,It$XL9E1%fDI.HM|$XI9IH;YHHfMtI.uL芃I|$X0HIHHIHII/tgHȇ'I9FFINHAH-HKHE~MII@蛇HWI&fDLfDLHԂI9M|$XHD$I$HD$I$I$HI$HLHHI$uLHD$耂HD$HtHmuHHD$bHD$H([]A\A]A^A_fDI$NL%T'I$^L"HD$HD$ HHLHHHD$HL$IHHzHH9H迁H/))'H)LpLh)H )H=|5)3I$1Hӱ})c)%E1HI)HLS)9)E1E1H)oHt9Hu#E~AFII 1E1LILIÈH ))*E1H)H))NE1E11Hg)0H=E1蹰H\)B)/H+)LLH ).)H=5) 21HZ))$H)릐UHSHHH5)HH9tAHXHt,LBM~H;rt&1 fH;ttHL9u1E1jtLHDDClHHHDLDHt H[]H$H11)M)H)H ))H=5)0H$fAWIAVAUIATUSHHHL%'H5K)HLd$ Ht$(Ld$0I@HE~+H)HgI@(HD$Ih Mp@HGHLd$MH=)Hl$ HWH;E'H;'OHOQE1 HYuLH QP H]';LHH́x OH H5'BR9:HHH'HEH52)HHHHHrHmH5)H9ZH-'H9kH9n{ ~ HCH;FH5Z)H9)H9kH9n{ ~ HCH;F3H=l)L9HCiHʄHH^ H=)H*HI Hm Lp-ImHT)A:)~1E1H)|MLDHH$IH''H7H 'H5 L rHII?H,H8I1A+H))H)H ݫH=-HH1[]A\A]A^A_L9uZL9u @IH{HHH;:'H;~'H)uHωT$zT$IIu LD$LH='HEHT)HIhHmuHzfH+uHzHHL[]A\A]A^A_HKHVH9DK DF DD@9A EH{0LSHA@IDA "LN0HHA@LD΃'*7A 9AHt&LH|1҅MIu LD$LH=|'I$L()HII,$Lqy@H`yc}@$f.Ld$fDHLd$L9upL9ucHyHHt=H; }'H;|'H)uHωT$xT$H )>)R1E1H)LE1bHZH ))H=5){*fDHKHVH9tHt HBDK DF DD@9A H{0LSHA@IDA LN0HHA@LD΃7A 9HLHz1҅H'HL9bHHL$|HL$LfDHt$ #HL9SHHL${|HL$=LNH2H{H7A 4LNHH{H7A 7A 7A HוH%MpIt)ItII@(HD$0I@ HD$(I@HD$ HtIHIMTHHD$0Lt$ Hl$(HD$HHt$8tHt$Hd*)<).1E1H)RH8);)!H)HE11H ))H=5)'}H zHfuHw'H5H8Xv1Hu);[)HD)rH=8xN1@OHt$ kHmHH5)HHVyHtHD$ HHH5))HHVyHtHD$(HHXH5)HHVcyHHD$0H%HŤ)?q)_1MHU)H_)=E);H.)H9HHxHHHt$rHt$HHt$qHt$HHt$qHt$HT$ LH5Y)LH-?H֣))Hf)LIsH)|MLDHH$IHGs'HWH GH5*L HII?HSH8I1AKwH$))P&P&H)H H=&!HH1[]A\A]A^A_L9uZL9u @IHoHHH;Zs'H;r'H)uHωT$oT$IIu LD$LH=s'HEHt)HIhHmuHnfH+uHnHHL[]A\A]A^A_HKHVH9DK DF DD@9A EH{0LSHA@IDA "LN0HHA@LD΃'*7A 9AHt&LHp1҅MIu LD$LH=r'I$LH)HII,$Lm@Hmq@$f.Ld$fDHLd$L9upL9ucHmHHt=H;*q'H;p'H)uHωT$lT$H-))&1E1H)LE1HzH ))H=5)fDHKHVH9tHt HBDK DF DD@9A H{0LSHA@IDA LN0HHA@LD΃7A 9HLHn1҅H'HL9bHHL$pHL$LfDHt$ HL9SHHL$pHL$=LNH2H{H7A 4LNHH{H7A 7A 7A HH%MpIt)ItII@(HD$0I@ HD$(I@HD$ HhIHIMTHHD$0Lt$ Hl$(HD$HHt$XhHt$HJ)0)|&1E1H)RHX))o&H)HE11H ))H=G5)qH %nHfuHl'H5H8xj1Hϙ){)m&Hd)rH=XlN1@OHt$ HmHH5)HHVmHtHD$ HHH5I)HHVmHtHD$(HHXH5)HHVmHHD$0H%H))&1MHu)H)e)&HN)H9HhlHHHt$3fHt$HHt$fHt$HHt$eHt$HT$ L.H5pN)LHM?H))<&<&H)LigH\gHv)\)&HE)HO)5)&E1H)YfDAWAVAUATUSH8HGH|$t$,HD$H1dHI Hպ)HD$HH@H;k' HD$HD$1HIHcD$,E1E1ME1ID$(LD,$HHD$ L;-9)fIuH;5i'tH~`HVHH$IvҺDIF $HHH9H9IEIFL,HIvLMIH|$Hj'I9D$xI;l$ID$L,HIEMtI(uLyeIHtH+uHbeIGH;i'H5`)IOHyHHAGHxH)HHDhHHI/LHL$dL;-)HL$t$(H|$H $iHH $VH\$ 1H $H)HIbHIH $# E1H~)fDH)HICH)JII9uLLHL$L$YgHL$HL$q HHsHH0uHL\$H $dH $L\$ID$($HHILHL$cHL$DH=9)H;=g'LMMHFH $kH $IMH=)LHL$L$HHL$HL$eI+uLH $JcH $HH $jHmH $u H'cH $HB)()TzMIE1H)HLL$L$L\$1LL\$L1H ))H=75)H|$L$LL$=Mt I/uLLL$L$tbLL$L$MtI)uLL$TbL$MtI(uL=bH8H[]A\A]A^A_I~ IFH5I)IIL9wH9nHD$(I~$H4HIF fD,$ILMI,$H+IGHHc\$,H)LL$L$HI_HHL$LL$=1H~*H)HHQH5r)H4HH9uHLLL$LD$H $,dHH $LD$LL$ H(u$HHL$LL$L$`L$LL$HL$H)EH d'HLHL$LL$L$G_HIL$LL$HL$UHL$LL$L$edHHL$LL$HL$HHL` HT$HH$HHHHLL$L$&`L$LL$cHIeHD$L|$HIGHLL$L$_LL$L$LLL$L$_LL$L$HLL$L$LbHHL$LL$H))zE1E1Ho)bf.HLL$L$?_LL$L$`I;l$VMlHIELD$LHD$HILD$gD,$LD$H$LMkcHL $LD$Hc'H2H9pLL$L$bL$LL$f.oH;7a'L]HZHގL)Mh)yE1HN)HMI8HHAAGHH HLLHL$aHL$H[L)M)yzMH)IE1H;b'H $aH $IH@`LHnAGAOHH H)JAXG`H7HLL$L$uL$LL$nHC)))yE1HH )Hl))OzMIH)H?))MzMIH)H))`yE1E11H)E1E1HD$gHԌL{)M^)yMHD)IE11HH).)zE1E1H)Hq))zE1E1H)HD))yE1E11H)E1E1H))zE1E11H)sHL)Mj)yMHP)I@HW)=)zE1H#)H;]'fH|$aHIH@HHHD$HBH;L)M)yMH)ILH $^H $D$($Hۊ)k)zE1HQ)DH[)A)yHE1E1H!)E1H~())yE1HE1H)E1HKL)M)zMH)IE1AWAVAUIATUSHHHT$iHIGhQwlLdHHH;I]']HxHHhL` HEI$H(I/L=0]'H\'L9H9UIEHt$LPHITL9I9SIELLH@pHH@HHH<IELHLPHIH+I/fD1HmuH7XMtIHPHIuLXI$HPHI$tZH[]A\A]A^A_fHT$LLP HH!HE1HHHxHWkLWfDH;-['H\H))lE11E1Hi)H߻,L$H M)g)H= 5J)E LWHWHoH@HbIML;=Z'IEHT$LLPHH%I/tH+YLVfDLV3HpVL;5YZ'L[Hj)P)lE11H4)HGH5 HPHX'H81x]H'Y'H5H8VHM))lH)HM1H ))H={5)`Hfx^Hޅ)n)lHW)Hd)J)lH3)hH5w)H=`)1iHHH%H+H` ))l1H) EH.H*)) mE1E1H)/ IH))l1E1Hd)HGTWHa)G)lE1H-)H7))l1H)HW'H5H81R[dH5)IyH*))l1H)3H))lHz) Hڃ)j)lHS)H])C)|lH,)aAWAVAUIATIUHSHHV'H5X)HHD$`H\$hHt$pH\$xXID$H~.HHrMt$0Ml$(ID$ HD$P@HNIIH\$PID$HD$HH=)Ll$`HWH;zV'<H;=W'HOQ E1 LyuLg*VH QP HU';< LLAIVx wp HiU'AI9MMMIEH5f)LHHIM#ImJH5ò)L9bL-;V'M9oL9nA ~  IGH;FH5)L9|M9oL9nA s ~ J IGH;FH=n)H9 IG LXHH H=)HHHj H+P HHm, H)m)V(11HR)ID$HH$L ɃHHkH [H5>LNH;S'HOH8LDRH1[WH4))''Hľ)H H=1HĈ[]A\A]A^A_I9uOH9uBLOHHH;fS'H;R'H.uHT$XOT$XHHJ)Hu LL$HAH=Q'HHD$(HH)HD$H)Lt$@D$8HT$0HT$D$ D$H$HT$P)HHH+u HuNDI/uLbNHf.IHVH9EW DN DD@DA9@A \I0M_HA@IDA 9LV0HHA@LDA>AADA2A9Ht LHP1҅fLHu LD$HHT$PH=P'IELLt$$L Y)Ӻ)HHJImLLHt$XGHt$XQHy))'11Hw)Hy)g)'HP)L11H 8)R)H=355)0QImMH*E1HCy ))'Hط)H=ςZLÀ(Ht$`IMH5)LIHVPMHHD$`%MH5)LHV&MHtHD$hIMeH5)LHVLHHD$pIH^x$) )7(1LH){H2x)޶) (HǶ)TL9LKHHHt$XEHt$XLHt$XEHt$XsHHt$XnEHt$XH4G@H#GHtw:) )Q(H )HMw))O(1H)mLHDI'H5H8GHT$`L@}H5-)LL2tHv))''Hk)fDAWHAVAUIATIUSHHH*J'H5˙)H-)HH\$ Ht$(Hl$0H\$8 HPH H .HcHfH\$Hh(Lp H@HD$H=C)Lt$ HWH;I'e H;J't HOQ E1 LyuLoIx WP HH'; LLAIbIx WP HH'AI9wM MMU IFH5ǡ)LHH IM I.H5%)L9L5I'M9wL9vA * ~ IGH;FH5)L9M9wL9vA ~ IGH;FH=)H9 IG L^LHH=H=K)HCHHH+HHmpH#t)ϲ)11H)I9uH9u @LCHH H;zG'H;F'H.uHT$0CT$ H5)H9$HG'H9EH9F}  ~ b HEH;F'M$It$ LD$HT$H=9G'IL<)HH I.L~BfH\$fDIH\$H\$fDIH\$hIHVH9wEW DN DD@DA9A CI0M_HA@IDA  LV0HHA@LDAIA|DA2A9oHt(LHD1҅f.AH5)H9L9uL9v\T} E ~  HEH;FM$It$ LD$HT$H=D'IL)HHcHLq))1LHܯ)L@D@${fHMHVH9DM DF DD@9A H}0HHA@HDA LN0HHA@LD΃7A 9MHt&L1HB!I$It$ LD$HT$H=*D'HH])HH H+u H?I/uL?HHH[]A\A]A^A_HMHVH9DM DF DD@9A H}0HHA@HDA |LN0HHA@LD΃7A 9Ht&LHA1҅I$It$ LD$HT$H=A'HH5)HHHn))H)f.H9uH9uH>HHt6H;JB'H;-A'%HmuH> HTn))211H)H1諺H裺H ̬))H=5ɬ)I9u ҐSH9u CL=HHt=H;~A'H;@'H.uHT$4=T$HmG)-)r11H)(DH9uH9u @H]=HHt>H;@'H;T@'HmuHT$)p1H%)LL$H|$@HD$0ؤLL$HD$@L¤L躤H|$P谤IGXH5})HIGXHD$0IG`IG`HD$@IGhIGhHD$PHIH|$0H9AI,$oHL$PHT$@LHt$0E1ٲEHD$0HD$@HD$PuLIHL$LHT$H|$0iH{W%) )p1H)H ))H=y5)HL$0HT$@Ht$PLGH5v)H=)1HIt@HI.uLl&HV)q)pE1HW)HVa)G)pH0)HV:) )ypH )L%HWAiHLV)ܔ)qpHŔ)H%Vϔ))oH|$0E1E1H)i- HU)y)o1E1H])D$ D$I1LHl,HHD$0-HU-))"p1E1E1H)HTU))pH|$0Hȓ)H=^LL$ HL$=(HL$LL$ YHD$0o)HuH&'H5^H8]%D$ D$E1YKIHTK)1)GpH)HzT$) )oH|$0E1H)fDHHHtHHt\HÐHuHTŒ))ڂH)1]H ))H=Q|5)~1HSy)_)‚HH)fDAWAVAUATUSHH8H-''HHl$ HFHHsH^H5()HFHCHHHFH1H8HWHHuH"HCH5)HHHIMH5EUL&I<$HWHI$.dH=)H\$HWH;W&'H;''HOQE1 La&H-x%'x W;UP HLAH%x UOH Br9HHVH8H[]A\A]A^A_LHRH RHH$L SH5`HIHHHTLBH XLHHw$'H81(HQL)2)WWH)H _QH=xzH81[]A\A]A^A_H$'H5yH8!H Q)̏)H)1~ILhXLp`ILxhH@XH@`H@hMtIEMtIMtILLLLի!H=y&LLLLHD$詫HD$HHHD$(HD$H8HWHHuHH5;)H{H9HXHHPH~)H;p1H;tHH9uH)H=.m)H9xuH m)HHH-l)H\H!'H9EuHHIMSHmH=v)Ld$ HWH;.#'NH;#'HOQE1 HYuLo"H-O"'x W;UP 6LLH"x UOH BR9HHI,$L[fDHYRDLo+LD$D$HC@^f.H?!@$f!@$2Ht$ 3H$Ht$H %DH})})p1E1E1H`})E1HD$1E10H>Y})?})p1E1E1H })E11HL$HHL$uHs'H5HH8HL$HL>|)|)PqE1E1HD$H|)1H|$1LHHHH=|)|)q1E1E1Hj|)1IH=\GHDuH'H5PGH8H=5|)|)EqE11H{)1HLHL$(HHHL$H7={){)>qE11H{)H1HALL$H_H=FHL$ HL$cH<z{)`{)5qE11HD{)%ffffff.HHHHtHHЅt4H'HHÐHteHHQHHuH HB<z)z)Hz)1脈H z)z)H=pe5z)襽1H;z)z))Hoz)ffff.AWAVAUAATUHSHH HEHA H1HHIMIc HHhHL-HII,$nH+THEHEH;L 'H;'L,$H;'xHUBE1 LbuL}uH 'H Q;P LLAILH H BR9 MHImM0H+LI,$HP:x)x)NHx)H x)x)H=@5x)躻HmuH DH[]A\A]A^A_H` LP L@ :H0 LXI,$BL 5LH)(ID @$f.HH`IH=9w)w)IHw)fDHH+HH8w)w)5Hnw)I,$uLO11 I7H8[w)Aw)7H*w)H 'H5^H8 Hw8!w)w)3Hv)"L}M(H]IHHmtILLH^I/ILzHHxIcHj HtE1GH=_A H 'H5bAH8*DAW1IAVAUATUSHH$(H$H$8HH '8H$(Hs)HxH9HXHHJH~(H;Z1fDH;\HH9uH$(LhHHpxHLH$@Hc@lL$HvL$H$@MH{L H{PLH HLL| LH$(HDŽ$0QH+WH$(H; '2H$(HHHHH$0H$IGH; 'H$IHDŽ$1IDŽ$DŽ$1DŽ$H$@H+H; ' IFHPHHHA^,IMePIEH9B$e Hc$IHĠHx$$H$H(H$f$IL$H$# Hm 'I9G H$I;OA IGL4IH$HtH+uH$$IFHP`HL;5'ZHH5Z)LH IM H-'HL'I9I9PUI$HDŽ$HHI$IFH5QZ)LHH IMm I9I9,II$HDŽ$HHI$xIFH5Y)LHH IM I9I9=I$1HHI$nIFH5Y)LHHR HH HmGIFH5LY)LHHi IM I,$ IFH5%Y)LHH H HE1H;'HQAHHIEEIMMPH$uH;-'E1DII?HEus$H=n)H5\L$ D$L$H$D$D$H$L$D$L$ 6 H;-'QH$H$H9$dE4H$HH$IL;%'EuH$H+$HHHHHPHH9HE1Hc$IHHI$H HԠH`xHc$HĠHɋ$I$$$1I9HDŽ$HH$H$H$H9HNH$ DEH$HH$H$IL;%'uHDŽ$DHc$$HDŽ HDŽ`HDŽĠL;%9'LH/Jn)0n)|Hn)@L;%'LKHX/n)m)|Hm)FL;%'LH/m)m)|Hm)LLE1K{LC{H lm)m)H=55im)dH'I9D$ ID$HPHH AL$HH$I,$|LoHI'I9D$GID$HPH H AL$HH$I,$L{H'I9D$n ID$HPH, HHA\$I,$L(HLHD$D$H$1HDŽ$DHHHIMePIEH$H=\h)H53 H,k)@lk)}HUk)H Ik)ck)H=V5Fk)E1g)fL%H{PLH|HLLH$(HDŽ$0:LH$(HDŽ$0H$@H$I;GMtIH$[H'f)f)|Hmf)H'wf)]f)|E1HCf)9HH'@f)&f)|Hf)IIIHH'e)e)|He)3IVH'e)e)|He)H&e)}e)|E1Hce)Y41fWHHAD$H$AD$H$H $ Hm&e)Fd)~E1Hd)H d)d)H="P5d)ͧ&H&d)d)}Hd)LwHHH$~H$H$H9HWHHH3fDHHA^AFHH LHHt0HHMHHQHHUHKHIMePHIE1ULHHH$H$H$H9HWHHtHDgHDŽ$4H?H%A\$AD$HH DLHHHHMHHQHHUgHZHDŽ$ HtpHuUAD$H$AD$H$H $H#b)pb)e}IE1HSb)LVH$LdH$|H#3b)b)|Hb)tHb# b)a)|Ha)MH;#a)a)|Ha)&LH$LH$LHLHH#BH"da)Ja)%|E1IH-a)H"7a)a)A|E1Ha)uLHLHqH;&PLHIH@HHH$t\H*HxHH!`)z`){E1E1E1HZ`)HDŽ$0H!X`)>`) |E1E1H!`)H!+`)`)|E1E1H_)fH$(plH&E1E1E1H8H)!_)_){HDŽ$0H_)H _) _)D}IE1Hi_)H]&H'H5JIE1E1H81H K_) 1_)2}H_)H5.])HxHb  _)^)o}IE1H^)GAVH;5[)AUATUSHHsl蝈HHL5&L9HxHL`Lh I$IEH('L;%p&L;%&u\HCLHHHSHHR(HHI,$uLImuLH[]A\A]A^@HLHHuH:])]) lH])M+Imt^I/LHLfDL=LoLH;&L4$H;~&IT$BE1 LjuM|$kH-&H Q;UP LLAIAUH H BR9|MM @$HLtIH;5&H &H5DE1E1H8H+X) X)5HX)1mfLefL]fH X)X)H=D5X)~H}X)cX)BhHLX)1fH >X)XX)H=D5;X)61H.X) X)IE1E1HW)NHWMW)W)sE1HW)!LIHW)W)[E1HW)IHW)sW)]H\W)HHH|&HH5H81HCW))W)AhHW)H6&H5AE1E1H8HVW) V)PHV)&7@uLLIHLXIlH=|!E1PHuH&H5o!H87+fAWIAVAUATUHSHHhH &H :)HHD$0HD$8HD$@H&HL$HHD$PjHFHH-\Hc4HHY&HD$ HK0HC(HD$Lc HCHL$HD$HI$L;%}&7 H=T)HD$HD$0HWH;&uH;N&HOQ4E1 HiuLo;X SH&P ;pLHt$IH H BR9 MM<IEH5B)LHHHHHU)H=&)H9xH&)HHL5|&)MPHLI>HWHI0IEH5zD)LHHIML;5"&L;5&IHPHI?%H5D)H9H-&H9kH9n { ~ HCH;F H5cD)H9C H9kH9n  { i~ @HCH;FLH5ED)H9H9kH9nh`{ X~ l"HCH;F.H5C)H ma"HL$ H; &H; F&DM1MO HT$LH|$IL\$ L$bQ)HHL\$ HxR)H^R)E1L\$E1H5R)`DH&HD$ H&HD$I%HY&HD$ H&HD$H9&HD$ HC8HD$ H5s@)HIHVCHHD$0|HCHH$L HHH H5^!LOH[&HNH8LHt1xHQQ)P)PPHP)H *H=[=ٓHh1[]A\A]A^A_H;&uH;5&uHHHH;~&H;&DHmuH6EEFH|$ H;=8&H;=& DM1MO HT$LH|$IL4$N)HHH$HD$O)FO)^E1H|$HO)jH;&uH;5p&uHHHH;n&H;&I DH)uH'EEHD$ H;)&AH;&D2EM1MO HT$LH|$IL4$M)HHI.XHD$H;&HHEIImuL~HtH+uHkHtHmuHWHL$HHD$HHHuH6MtI,$uL"HhL[]A\A]A^A_L;5&L[HJHD$N)6M)E11HM)H|$LD$(D$(^LHD$H5&HHHIH;x&L;5&+IHQHI]H|$H7HVHt$HHbHyM)H=)H9x"H)HHL=)M~IGH5C)LHHSIMI/~HH&I9FHL蹋IMI.SIGH5H5)LHHIMI/=H5fL)LHII.L;= &L;=&DIHHHIE"HD$H5K&HH9pH|$HIMHL$HHD$HHHH_f.[@$efL;=1&\LAMHy L|$:K)U K)E1E1HK)H|$fHKHVH97DK DF DD@9{A  H{0LSHA@IDA  LN0HHA@LD΃  7A 9(Ht'LE1H)HHL\$ 9HT?)P?)E1L\$E1H>)LfD7A gH;-&^HARLNHH{H7A ,H54)HXE H|$ B MA1MO HT$LH|$IL4$*=)HHLH_HD$>)R>)6E1H|$H=)fD7A `LNHH{H\7A B7A u7A iHt$ILfI HCJcHHF8HD$PHC0HD$HHC(HD$@HC HD$8HCHD$0HqIIHCJcHH &HtH5(E1E1H81_H8HD$<)W<)H|$H<)H5 )H= :)1?HI H~I.uLsH<)7x<)Ha<)HHD$b<)6H<)1E1H,<)H|$IH=3)H;=&HCH(HD$H|$] Ht$H= 9){HIHL$HHD$HHHuHL}I.uLsH;)5x;)Ha;)HLt$f;)4L;)E1E1H/;)1H|$HlHD$);)4;)1E1H:)H|$H=b+)]IYH=N+)H )H5 )zI7H=7|H1:)2:)E1Hx:)CH12)P$2)H 2)HQ2)O1)H1)kLyD$ PHDD$ PH 1)J1)H1)%H1)T1)^E1H|$Hq1)3H{1)Ta1)kH|$HE1);IH|B1)T(1)hH|$H 1)I]D$ qHDD$ H,0)N0)H0)FH0)T0)nH|$H0)WHHt$(蓿Ht$(zH0)Gf0)uHO0)H=)H)H5%)I_Hq70)T0)fE1H|$H/)H=)mIDAWAVAUATIUSHHXH-)HH|$HD$0Hl$8+HFHHHH$L HHH H5LNH&HOH8LBH1HL/)2/)PUPUH/)H _H=rHX1[]A\A]A^A_DHn LkHI/)HEH  )H9HH )HhHH )HHCH5D)HHHIMH+I IE I,$8 H.)H B )H9HH! )HHL= )MIGH59$)LHHHH%I/> HW&H9CHD$LHlIMH+ H.)H y )H9HHX )HDHHD )HDID$H5)LHHIM"HCH;D$D$ D$E1H;&UH;& LL$(HILL$(>MtLHHcD$HEHIlHcD$ M|HCLML &H QA;P 1LL$ LHAHD$[H LL$ QP A=p 9 H|$I. H+ Hm ID$H5})LHH^IM]H5)L9 IBH;)&LIzuAz fH)&H-&E1HH9AAIHPHI; E HHPHH/ EH+)OH )H9HH)HyHH)H5HCH5_!)HHH IMH+ ID$H5)LHHlHHHCHt$HH@pHH@HIMzH+ ?HHLpZHIH*)H )H9HbH)H>HLk)MIBH5g)LLT$ HHLT$ IMI*R H5)LLLL$ LL$  I) IGHHE{L &H QA;P LL$ LHLIJH LL$ QP A=p4@9 MI/ H+ I.r H|$H5)HGHH8IMHD$I9GIoHMwHEII/ LHLHmHuHiH4I. H+ ID$H5)LHHIMQLL$ HILL$H( )LLLL$HIFH )HHILL$CI. I) Ht$LLLD$蜶LD$LLD$"HILD$ID$H@pHH@HLD$LLLD$HD$H|$ I.%I$LHl$i@H I)H9HH()HHH)H HCH5@)HHHrIMUH+H>&I9GHD$LLeHHI/H|$H5R)HGHHJIMNHD$I9GM MgM@ MoI$IEI/HLLn I,$IuLLT$LT$MIm^I*uLHH\$E1E1HHHHHuHLD$蒶LD$MtImuLLD$tLD$MtI(uL\HmuHMHD$HX[]A\A]A^A_fLL13I,$HLHo%)@H )H9HH)HzHL)M7IBH5p)LLT$ HHLT$ IMI*IFH;D$0D$ 1E1H;f&H;)&LT$(HILT$(MtLPHcD$ I$HMdIEMlIFHH6L R&H QA;P 1LL$ LLH豸H LL$ QP A=p4@9HCI/}I.H9H;&HHPHHH#)H )H9HH)HHL5)MIFH5>)LHHIMI.HD$I9GJLLaHH I/I,$H|$H5\ )HGHHWIMHD$I9GAIoH4MgHEI$I/HHLx HmIuHLT$LT$M I,$I*uLʲHH\$E1Hl$E1HH4@qDHA3H!) !)>VHs!)Hl$Ld$E1E1E1E1E1E1HLD$LL$ LT$ /LT$L.L.LL$ L.L.H !)"!)H= 5!)dH|$LD$BH\$HD$HfL許5H蘱H舱LxLLT$cLT$fLH7H8:H(LHHC@LL$0Hl$8Ht0L|$@HSE1HBR uLSLL$ HLLL$ HD$H|$\Ln-I/vL|iLhUk@$.fH;A&H蛵HP)6){VH)f.LLD$LI$Hl$LD$2DH-&AHEHJfDL谯@H蠯LH舯VLxHh LLL$ SLL$ fL8LLT$#LT$AfIF@Ht0HLLT$0LT$ Ld$8Ll$@I HHLT$ 'L+VL谮'H蠮OL萮5L耮LpH`gLPMS@$)f.LLD$#LD$hfLLL$HD$LL$LD$4Lv@$Rf.L踭L設+L蘭Ht0HHLL$0LL$ Hl$8L|$@YHHD$LL$ Hx) ^)VHG)E1E1Ld$E1E1Ht0HLLT$0LT$ Ld$8Ll$@YHHLT$ #HD ))]VH)E1E1Ld$Hl$E1E1nfDH))VLd$Hl$E1H)E1E1.HnHBH/HHL蒪HIHu*M~.H5$)LHVHHD$8IMLl$0Hl$8H=)H0(H51(H4H%))xUE1E1E1H)E1E1E1HD$J蕳IH))vUHk)E1E1E1HD$E1E1E1H=)HHF HD$8HCHD$0LLMRYHHLM)VE1MH!)E11Hl$E1E1HU) )ULd$H)E1E1E1E1E1MgM%MwI$II/LLLHI$MHHI$LlrI3Hy)_)UE1E1E1H?)E1E1HD$H=)HH`&)  )UE1E1E1H)Ld$E1E1E1H=)H(H5(~HSH))WE1Ld$Hl$H)E1E11E1(H)m)kVLd$Hl$E1HI)1E1E1HK)1)WE1E1Ld$H)Hl$E1E1rIvH;))VH)E1E1E1Ld$Hl$E1E1bH5+(LIHVkHHD$0MHCH)y)KVLd$Hl$E1HU)E11E1E1HT):)VH\$Hl$E1H)9H|$IH=LL$ ~LL$ H*))vVH)Ld$Hl$1LLHHt葫LT$ IH)y)IVE11E1HZ)H=(I}H=(H(H5(I[Hh.))@WE1E11H)Ld$E1Hl$LLLD$KLD$HD$yH ))>WLd$Hl$E1H)E1E118H)})4WLd$Hl$E1HY)E1E1E11DH跦HMH&H5@H82H\"))VLd$Hl$H)ݩIH))VE1E1E1H)H=N(H~H=:(H(H5 (6H\H;&ZL׺LT$ СLT$ HHHN) 4);VE1E1H)[H=(H$(H5-(HH9MH\$))UE1H)E11E1E1cH))UH)臨IH) t)9VE11E1HU)H=(HHMVMxI^IHI.0HCID$ 1ҿHB))VH)Ld$Hl$E1E1E1E1lH))VE1Ld$Hl$H)E15H)z)VHc)Hp)V)VE1E1E1H6)Ld$E1Hl$LT$ I=H[!))VE1E1Ld$H )Hl$E1H=(LIH=m(H(H5(iIH|$蘦I HH )w )VH` )H=TLL$ ڡLL$ LHLmHI H\LXDIH< ) )VLd$Hl$E1H )E1E11E1E1aH ) )eVH )H&H5H8dH=(IYH=(H(H5(I7H-B&H9H;&DAA H;&A L׾LT$ LT$ HLLT$(IɜHCD$ 1ҿLT$(bD$ 1IHLL$(L~IFD$ 1ҿD$LL$(gD$ D$HHY ) ? )VH( )H1LLL$ HD$XHT$0LpH5w(HLH- ) )@U@UH )H ) )VH )H ) )VHo )IoHMoHEIEI/uL1LHLHHEMHHHErHeff.AWAVAUIATUSHHHGH5(HH8HHHCH;K&HkHrL{HEIH+H &IGHl$H9HL$#H;T&IWBE1 HZhCL5&H QA;P NHLIH Aqp B9MPHmuHϙMGI/H=(H5(HGHHGHHH=)Ll$ HWH;T$H;b&HOQ|E1 HYDOL5&H QA;P LLH&H Aqp B9HIMwHEH;s&iLmM\LuIEIHm(LLL ImHuL舘I/HXI.HLHHaI,$H+uH;HHH[]A\A]A^A_fR9@$@LR9@$@MgL踗H託"L蘗L舗HxLhHt$ CIbfHt$LCIHH G&H9HL$H;&HSBE1 HjuLcL5b&H QA;P v1LIɚAH H BR9|M&I苚@$11HBIH;D$L|$0H;E&(HUBE1 HZuLm4L5&H QA;P LLH Ap p BR9|HIAf˙@$Ht$0H3BHL躴IHL觴IH=AH)l)1HS)LHH1 1H +)E)H=5()#GHW))1H)땀1Ht$ 0IH))1E1H)I藜HH))L1Hh)^HјHH&H5ZH8"Hv<)")LH)HL))H)H=`E1Ht$LIHuH&H5H8nHE1H=[vH5)1HIuoLH8H;H=蠖1 {HuHg&H5H8ГHL&H5H8赓!Ht$0HHfDAWAVIAUATUHSHHHg&HHDŽ$HDŽ$H$:#HFHHAHF(HD$(I^ MvH) HD$PHD$`HD$pL(hE111AHHAHHD$D!HHHHD$DxHA)L(hE111AHLAHHD$  HHH6EHD$ P H萖f.D$p IFIFHPHCHHcHM*"LHH$H5H&H H5L iHOH81LBH#H( (44H(H H= B1HĨ[]A\A]A^A_HA&HD$(H5(H|$(_H5(H|$ (HD$(H;& Hf(H(H9X8H(HHL5(MzIFH5(LHHHHD$PI. H(H+(H9XH (H#HL5(M#IFH52(LHH7HH6I.' HD$PH=&L@H|$0I9zAE1E1L;&L;&N H衒HHD$`$MtLpH|$(IcIcHHHH|1H|$PH\HI-$H|$`H/ HD$`H|$PH/ L|$PIH Ld$P1HD$PH(At$I|$ L(IHL$ 1HT$LAH<H;Ƒ&HD$`L-&HtH+uH詍HL=K(HD$`HQLHL$(HHT$HHlH@HT$HL$(HH# HHHHLLL=(IPLLD$(HHT$脒HNHHHT$LD$(LMLHAHHD$P4HHH;L$0HD$p HPHHT$p H@HHH|$PHD$PH/q Ht$pH|$PH :H|$pHD$`HtH/uZHD$`HHD$pp"H|$PH/ H|$`HD$PH/q HD$`E1HD$HEHMHHE Ld$(H\$0MHMH@I0HHL0I@H0I8H0H6蜒IEAGE1IG ,xfDH(H0H0H@(AE;O}NIcI4H0H@H0Pt8H(AHR8HcR H0E;O|IM9)Ld$(H\$0MH|$ʉH5S(1H9H+HHHmI$HI$LHHI$uL蔊MtImuL耊H|$tH|$HHD$HHHuZH|$ HtHHD$HHHu7HHcHH|$`HD$PH/vHD$`~HD$8HDHL8Mt L;=&HPHuLhH@MHD$0tIMtIEHD$0HtHHUHH} HD$[HCHIHD$`7HD$8HL$0LLHD$`H H5(1LlI.HHH+LlH }(1HT$ Ht$HkH;&HD$PL-&H(H6(HD$PH9XH (HHL=(MmH5(LHHD$`!I/ H51(LHIH(H(H9XHn(HbHHZ(H!H5(H2HIH+H|$`HP&LOHD$0I9FA11L; &L; n&H׉t$THHD$pt$HtHXHcH|$`HHL|1NtHHD$PH|$pH/HD$pH|$`H/HD$PHD$`HH|$PH/vLd$`HD$PLHD$`Ht$(L$H$H$H<H)HI}Hج(b (b5E1E1Hg(H1o(U((YH>( H 2(L(H=5/(*.HcHD$P (d (5H(1E1E1E1E1H%(k (61E1E1H(aH=V(H_(H5h(RIHԫ(e (5E1Hf(LOEIAAHH$L$L$HHD$Pt-HI/uLzI.LzH8(e (5E1H(yHH$L$L$ 'HHD$PvH۪(e (5E1Hm(H_HtAHGHHH|$`HD$`H/u.zH|$`A1LOA1mHy.HC (e (5E1H(H(e (5E1E1H(WH=L(HH=8(H(H5(4HsH|(e b(51E1E1HC(L&yHz@(e &(51E1E1H(HK(e (51E1E1H(H=|(GIGH=h(H1(H5:(dI%H5(Hx觞Ll$Pfx[xfDKxRfD;x*HU(e ;(51E1E1H(E1H]#({  (6H(I.LwH]wfDH(d (5H(w6Hӧ({ (6Hh(qHr({ X(61E1E1H9(E1Hz@(b &(k5E1E1H (1AIH@(~ (71E1E1H(|LhH@HD$0H( ($8E1E1E1H(:HϦ( {(7Hd(I.ta1E1E1E1H\HD$`Hz&LH8JvHD$pHj0( (7H(L1E1uE1E1H~z&LH8uH( (71E1E1H(E1UH(~ (71E1E1Hw(E1#j}IHq(~ W(7H@(LH=(诛$H=(H<(H5E(ϛHPHHT$PcH@HHH|$`HD$`H/utHD$`A1ɺAHx4H(~ (71E1E1H(E19HΤ(~ z(71E1E1H[(E11AHX(~ >(|7H'(3|Ha'(~  (z7H(H=(eSH)(x (6H(H(w (6H(H=;(HĹ(H5͹(7I$1wHGHi(v O(6H8(DH5(Hx=Ll$`EHw&LH8sHH(k (61E1E1H(H(_ (,5H(H1((WYIE1H(1E1EH n((H=5k(f$HHD$`\(j B(6H+(HoI2(j (61E1H(E1H=(` (55H(Hfv&LH8qHD$PH(k (6H(H+31E1E1-H="(H (H5(IL蟚H*HV(b <(I51E1E1H(E1H=(艗I7HJ([ (5E11E1H(E1E1AuH|H(u (6H(Hڠ(Y (4E11E1Hg(HD$ E1E1HE116pE1H=(貖IHs9(b (N51E1E1H(E1wHH4(b (P5E1E1E1H(owH(b (K51E1E1H(E13LpMtBH@IHH|$PHD$PH/uEoHD$PA1AL@QAE19LfIt0ItIfHF(H$IF H$IFH$HmIIIMtxM"H$L$H$HD$(^H$LsH5%X(LHyHu( [(}4}4H?(H5(HHVrHH$IH5(HHVrHH$t:IM8H5~(HHVrH>H$IH=MH( (s4s4Hv(IFkHPm_Hj({ P(Y7E1E1E1H0(E1Lm H|$PH_%(|  (7H(H|$`HD$PH|$pHD$`H ((H=5(HD$pH|$8HL$pHT$PHt$`HL$pHT$P1Ht$`uH1HLHD$@I.HHL$@uLHD$HlHT$HHL$@H)uHHT$@lHT$@HHHT$@gHT$@AHHHHH uHkEH|$`H|$PHD$`|H|$pHD$PiHD$8HL$0LLHD$pHdH~({ d("7HM(HD$8HL$0LL1E1E1E1HH\$pLt$PHl$` oHHLH|HEHD$`HD$PHD$p({ H((.7mH({ (&7H(FHܚ({ (7Hq(H{({ a(7HJ(HT(b :(}51E1E1H(E1H\"(b (r5E1E1E1H(H,(k (#6H(,fff.AWAVAUATIUSHHjHMZH{H@t}HiHҙ|(b(HA(H H=>HHPHHuHigH[]A\A]A^A_@|@ts1lHIc1HHpImItdMCIL$@tWLHhI$HPHI$LjhDHk&H5H8JiL=hHtk&H5%HH81o1HLLHHHHI;HHj&H9CLHI"E1ILImt]MtbI,$tGLI.uLgH((H(9LbgLXgI,$HƗp(V(H?(tfDLgH (3(5(L{M LcII$H+uHfLLL2IH/((H(XfAWAVAUIATLgUHoPS1HH@lADpAAAE~kfDHt)HcHHTLH:H|H2H0II4HIݐx?H=h(H5ILt7HA9H[]A\A]A^A_f.I͐xH%((GH(fH ((H=5(c1ffff.SHHGHHOxHHH|$HD$Gl5HLLD$ LLD$`LL$LALD$(LBLD$hLFL$LALD$0LBLD$pLFL$LALD$8LBLD$xLFL$LA LD$@LB L$LF L$~mLA(LD$HLB(L$LF(L$~GLA0LD$PLB0L$LF0L$~!HA8HD$XHB8H$HF8H$Ht$HHH;g&H9VHHHHHHHHuHHD$cHD$H[f.H1Ht$ H2HDŽ$Ht$`cHqHt$(HrHDŽ$Ht$hGtHH`%1H$`{$H$h$~sH$p`$~_H$xB$~KH$L$~7H$.$~#H$.$~H$"1H99D@CD$D$)؅HcH`HcH`H̠HHԠHKHcH`HcH`H̠HHԠHKHcH`HcH`H̠HHԠHKEHcH`HcH`H̠HHԠHKHcH`HcH`H̠HHԠHKHcH`HcH`H̠HHԠHKtzHcH`HcH`H̠HHԠHKt9HcH`HcH`H̠HHԠHHDŽ$`H$HDŽ$hH$HDŽ$pH$~xHDŽ$xH$~_HDŽ$H$~FHDŽ$H$~-HDŽ$H$~HDŽ$H$HH$H]9L݅!E1E1L$ J@JDŽAJxH=<(H5pP!ID9zJ`OtDL9tHt\IcAo^HI#L^HI"H^HH!1_HH{!H=λ(HX(HLhLp jcHH"HmH=l(H\$PHWH;_&6H;_& HOQ"E1 HiuLw^L-5^&p VA;UP !HLH^p AUNH BR9$H)!H.!H+Hp HmHT(:(H#(H (6(H=ڑ5( DlXHa ((H([H ((H=ٶ5( XH[]A\A]A^A_fDHpYhH`YH HmHEY K]@$fH$`DH$XH$HzHHuHHЃiH$hH$HDHHrHHȃ3H$pH$HHH0HHȃH$xH$HHHHHȃH$H$HHHHHȃH$H$HlHHHHȃ~_H$H$H:HHtHHȃ~-H$H$HHHHVHHt$8L0H$0H$(H$pHH(HFHHȃiH$8H$xHHHEHH3H$@H$HVHHHHH$HH$H HHHHH$PH$HHHHHH$XH$HHH=HH~_H$`H$HHHHH~-H$hH$HPHHHHHD$8H9XHD$@H$PH$H$PHHP`CHHĐH;PHHKHcH̐HPH9HHqKHcH̐HPH9vHH6KHcH̐LHPH9;HHKHcH̐HPH9HHKHcH̐HPH9HKHcH̐H9PH~WKHcH̐mHPH9\H~#KHcH̐9H9P+H$H$ HH$ H`HI`H; HCHH`H H9pHmCHH`JH H99H6CHH`H9 HHCHH`H H9HHCHH`H H9HHCHH`hH H9WHH~UCHH`2H H9!H~"CHH`H9 D$Ht?L$(RH$pH$0E1LOH$PHP`HcH$`H`H9H$hHL$aH)HD$HȹHI9L@HIFHf~؉)HcH0HcH0HpH̰HpH԰MHcH0HcH0HpH̰HpH԰MHcH0HcH0HpH̰HpH԰MEHcH0HcH0HpH̰HpH԰MHcH0HcH0HpH̰HpH԰MHcH0HcH0HpH̰HpH԰MtzHcH0HcH0HpH̰HpH԰Mt9HcH0HcH0HpH̰HpH԰SHDŽ$0H$pHDŽ$8H$xHDŽ$@H$~xHDŽ$HH$~_HDŽ$PH$~FHDŽ$XH$~-HDŽ$`H$~HDŽ$hH$HH$H(IgL9|$FH$PH$0H$PAHHD$ Lx`C ~lHcHL;DL1/@HcȉHH̀DH9HcH@9HuLcH$`L|$(J`IIH9 H$hHH$aH)HD$HHH9HpLHCHE~{H$I~PH8IIuxHEI~IOH$L LL$K8L$H$ILH)8MMHIE@L$L$H$L$LHHHޱHDD狔$UL$HmuHL$.4L$MtI/uLL$ 4L$L$L$LH$H$HHĨ[]A\A]A^A_1@H6&H5 E1H81%;E1E1tHd((8fE1HDŽ$Hy(H$H$4L,LE1!H$HH :(T(H=57(2E1E1HYfDH2vDIM`IExHcH H AT$HcH H AT$HcH H AT$trHcH H AT$tXHcH H AT$t>HcH H AT$t$HcH H AD$t HH EXEI~IOH HH$Y5H$IHm3E1IcHH$5HH:E1ImuLz1E1E1*Ha(t(*fH](H5&H5H822HaQ(7((f1E1H(11=4I Hja((&f1E1Hޟ(`H>a(Ο($f1E1E1H(1f.USHHHOHDHWxHHH|$HL$OlADž5HH:H|$ H8H|$`H>H$HzH|$(HxH|$hH~H$HzH|$0HxH|$pH~H$HzH|$8HxH|$xH~H$Hz H|$@Hx H$H~ H$~mHz(H|$HHx(H$H~(H$~GHz0H|$PHx0H$H~0H$~!HR8HT$XH@8H$HF8H$LC`H$Ht$H`hAXH$w3H|H$HHH$HHHĈ[]H2Ht$ H0HDŽ$Ht$`bHrHt$(HpHDŽ$Ht$h;HrHt$0HpHDŽ$Ht$pHrHt$8HpHDŽ$Ht$xHr Ht$@Hp HDŽ$H$Hr(Ht$HHp(HDŽ$H$Hr0Ht$PHp0HDŽ$H$oHR8HT$XH@8HDŽ$H$IH](g(!xHP(1H B(\(H=݊5?(:1aH]2((,xH(ffffff.Kff.ATUSHHHOHDHWxHHH|$HL$OlAᧅ5HH:H|$ H8H|$`H>H$HzH|$(HxH|$hH~H$HzH|$0HxH|$pH~H$HzH|$8HxH|$xH~H$Hz H|$@Hx H$H~ H$~mHz(H|$HHx(H$H~(H$~GHz0H|$PHx0H$H~0H$~!HR8HT$XH@8H$HF8H$LC`L$Hl$H]A8L$H/HoHLHHH@HHİ[]A\H2Ht$ H0HDŽ$Ht$`lHrHt$(HpHDŽ$Ht$hEHrHt$0HpHDŽ$Ht$pHrHt$8HpHDŽ$Ht$xHr Ht$@Hp HDŽ$H$Hr(Ht$HHp(HDŽ$H$Hr0Ht$PHp0HDŽ$H$yHR8HT$XH@8HDŽ$H$SHZɘ(|(wH(1aH ((H=M5(1dHYz(`(wHI([ff.AWIAVIAUATUHSHXH,&H9H$LHIH9YH$HHIIFH5:(LHHHHHBHBHHH HcHfDE1DH*HEH5Ä(HHHrHH=HBHBHHH HcH1fDH*&AH$LHHLHDHHHX[]A\A]A^A_DBJHH HcAH9u{AH$A+HH$HWC()(,nH(DBJHH HHcAH9tH$NH$ApDrAd@DrWjH$+HH$HV(s(0nH\(H$H M(g(H=05J(E1fDBJHH HHcȉH9bH$MH$FfDj8BJHH HcȉH9H%eH%HH$0*HcAH$H9H)HH$l@HH$)HcȉH$H9H)HH$fHH$&CHH$HH$H$|H$AH$H9HwHH1FHH$$H$)HT(~(:n1He(HH$BHH$HH$H$סH$H$H9HwHH1ZHH$#H$=HJT(ڒ(.nHÒ(b+HHT((!n1H(,H5(H}IHS(e( n1HL(HSV(<(n1H#(H57(Hz(IHkS((n1H(HBS(ґ(*nH(Z*HfAWAVAUATUHSHHv(L-v(H_&&HH|$HH\$pLl$xH$ HFH ~'H HsHF(HD$PLn H^HWH%&HD$PH( H(hE111AHHHI HH]HN( H(hE111AHLHH HH'EA;D$QIH&f.D$XL&f.H(\D$XHsf(H9PD$`HLf(HIHL8f(MIBH5D(LLT$hHHLT$hHHI* D$`T#HI`H!#&H9CLHLT$hLT$hII* MH+ L;=[$&L;=#&6 IHPHI RHD$HD$XLI"HHD$`"HIH(H=#&IHt(Ht$HALT$LT$XLH|$@D$8H HT$(HD$0D$ HD$D$H$HT$PH=j"&(HHD$HLT$XI/ H+IE1E1HHIuLI,$uLHtHmuHMtImuLMtI.uLHD$HM LHH$IH!&HOH OH5]L *RHII?HVH8I1A%HN(gh(H$H$HL(H NH=f|gD1HĘ[]A\A]A^A_H(HBc(H9PH!c(HwHL c(M<IBH59t(LLT$XHHLT$XHHI*dHCH;) &/ AE1E1H;!&pH;Q"&LT$X9!HILT$Xz MtLPHEIKlI$OdHCLM !L  &p VA;P  1LL$XLHAI H LL$XQP A=H @9Mo I/yH+wH (IL0HI HH;Hԋ(Hma(H9P HLa(H; HH 8a(H HAH5(HHL$XHH HL$XIM H)HX(H=`(H9x@ H`(H HL`(M IBH5{(LLT$XHH LT$XHHI*sHCH;& H;s&Ll$pH;1 &cHSBHr1ɨ Ht$XHL$`L &H QA;HL$`P ALL$`LHHD$XHD$XLL$`p ANHt$XH B9HHIHs I)H&I9G HLHL$XHL$XHH){HX I/XH;Q&H;&DHHPHHeE HL$HH=&MH6n(AHHq HHH|$@H(HT$(D$8D$ Ll$HD$0HD$D$H$HT$PH=&(HHD$HH+XH`KLPH@H;)&HAHqI7((%H(Ld$HE11E1E1LLL$PHL$`LT$X试LT$XL袕H蚕HL$`H荕LL$PL耕H (Ç(H=v5(H|$HtLd$HHD$HRfHQ&HD$P5H9&HD$P!L(LH|HLLHxLHL$XHL$X+fHHC@*LT$pHl$xE1L$HCHtpL@@ uLKLT$X HLALT$XIMY L@Lz@$Vf.L;=&L;H*G(օ($Ld$HE11H(1E1E1E1DLHx1LhLXCHLT$PCLT$P fLLT$P#LT$PfR]DHKHtpHHLT$pLT$XHl$xL$YHILT$X` MI*LHt$XHt$X@$Ht$pHHLfIt+ItIHF(H$HF HD$xHFHD$pH`IHIIiM H H$H\$pLl$xHD$PH$E(Ѓ({$1E1E1H(E1HD$HE11E11HD(($E1Hq(Ld$HE111E1E1E1aLCMLKIIH+uHLD$`LL$XLL$XLD$`LLLLD$`LL$XdjLD$`HLL$XI(LLHL$`LL$XHL$`LL$X+HDۂ((%E1Ld$H1H(HC((%1E1Ld$HHo(E1E1nLSM5LkIIEH+/IELA1ҿA1LHTHI]HQC((c%H(Ld$HE11E1E1E1HC݁(Á(X%H(1E1E1Ld$HE1H=@n(HW(H5W(<9H=HB(j({%1HQ(E1Ld$HE11E1FH&H&H5@LH8 H= LLL$XLL$XLT$XHXH'B(Ӏ(%Ld$HE11H(1H=Tm(8IH=@m(HiV(H5rV(<8IHA(j(%Ld$HE11HI(E1K<HL$XI-HxA>($(%H (H=l(|7HHH5n(HHVHtHD$pHHH5q(HHV~HtHD$xHHH50h(HHVTHqH$HjH5b(H=|(1襂HH+HaH+uH#H|@B(((%Ld$HE11H(1E1E1HC@ (~(%Ld$HE11H~(E1MGM}(}(6%H}(H=*j(4IH=j(HS(H5S(5I^H>Z}(@}(%H)}(xHm>3}(}(%Ld$HE11H|(E1E1H6>|(|( %H|(H5_(H=y(1HIKH轾I/,H=|(|($Hq|(Ld$HE111E1E1E1E1^H=c|(I|($1E1E1H*|(Ld$HE1E1$Hc=)|(|($H{(!LT$hHH*={({($H{(IH=ch(.3ILkMLsIEIH+LLLLT$hLbIIELT$hHHIELLT$h2 LT$hH<G{(-{($H{(?HLT$XL AIE1ҿALT$XcLH*H)z;HeH;z(z($Hz(H=4g(HP(H5P(02Iq,D$`HD$`H;Xz(>z($H'z(AE1HT$pLCH5'LH8H-;y(gy(4$4$Hy(lH=DLL$hHL$`2HL$`LL$h1Ht$pH{HLO H:iy(Oy($1E1E1H0y(Ld$HE11E1(Hg:-y(y(%Hx(HLT$h LT$h^H):x(x(R%Hx(1E1E1Ld$HE1H1LALT$XIO HH &H5CH8] H9wx(]x(J%HFx(@AWAVAUATIUSHHH &HH|$PHDŽ$HDŽ$HDŽ$H$HFH H`HF0HD$XI\$(Il$ Md$HAx(L(hE111AHLAHIOHH9 Hw(L(hE111AHHAHIHH Hw(L(hE111AHHAHI HH AEA;Fu A;GH8w(H 1O(H9HHO(HHHN(HHCH5l(HHHIMH+ H5&[(LHHID$H; &H;< &H$H; & IT$B Hz1 H|$H L R &x WA;P  HLL$`HHD$HH LL$`x AOH B 9H!LH+P HfH)- H;-f &H;- &9HEHPHHU) H=j(H;= &H5Y(HF0HH"H==r(HHH)#HmuHHҶH+uHH5t($ t(<Ht( ID$HH$H6H&H 6H5tDL 8HOH81LBH= Ho55t( t(::Hs(H H5H=Ic 1Hĸ[]A\A]A^A_H&HD$XID$ID$HHSHXzHcHHs(H5K(H9pHK(HHL%K(MMID$H5i(LHH!HHII,$ H5W(LpHIHm&H9CCLHHI,$ HH+ H;-&H;-&& HEHPHHUc JHr(H=J(H9xHlJ(HHH-XJ(HHEH5h(HHHIM|HmHar(H5I(H9p7HI(HHLI(MIBH51a(LLT$`HHLT$`HD$HH|$HAI*Hq(H YI(H9HH8I(HkHL $I(M$IAH5h(LLL$`HHLL$`IM,I)+I@H;&E11H;& H;&lLD$hLL$`HHLL$`LD$huMtLHIHLtIELlI@HH"HL$hLD$`TL &x LD$`HL$hWA;P 1HLL$xHL$pLLD$hHD$`p LL$xLT$`LD$hHL$pVP A=p4@9MyH)I( HD$HH@H;9&E11H;& H;c&LT$hLD$`FHHLD$`LT$hAMtL@HD$HHLTIL|H@HH;HL$`L &H QA;HL$`P 1HLL$hHL$`H|$HHx LL$hHL$`WP A=p4@9HH)Ht$HHHD$`HHHH&I9D$*HL蛭HHmHI,$H;B&H;&}HHPHH>HD$PMHT$XH=&HHp HHD$0HZZ(L|$ D$Ll$$HD$(H5[(L Z(HD$k(HHDH+uHiI.uL[MtImuLGMtI/uL4HW@L LL(H;-&HCH2.l(# l(;Hl(1E1E11E1E1HD$HNfH;&vHgH-l(' vl(<H_l(1E1E11E1E1HD$HH(HHHL$HHL$HfHHL\LLD$`LD$`fLLT$`LT$`fHx6HhLX*HHLI@@L$L$E1L$I@HLP@ uMXLD$hLL$`HLALD$hLL$`IMLLT$hLD$`xLD$`LT$hDHD$HH@@#HD$HL$E1L$L$HH@LH@ u HL$HLYLT$hLD$`HLALT$hLD$`HH<LLT$`wLT$`I*vLifDHLT$hLD$`LT$hLD$`HL$pLT$hLD$`HL$p@$LT$hLD$`@HHL$`HL$`@$H;-Q%HH*`i(# Fi( <H/i(cf.LHHLL)RGDIl$fD1fDHE>HEHHH voHcHDE1DHC7HCHHH JoHcHD1fDHɚ;KH=](H;=%H5L(dHFF:IM H=Ge(LLT$PHHD$HLT$PI*H\$HHЩHHD$PHHHH(g( g(b;Hvg(1E1E111E1E1HD$HH|$HHL$hLD$XLL$PLT$` uLT$`LuLtHtH1tLL$PLtLD$XLtHL$hHtH f(g(H=3V5f(M\IfAT$HcHHT$H"HHT$HH'f( f(+;Hrf(DDeAMcIHT$HHHT$HH'Kf( 1f(5;Hf(DDeEII IDDewDeEII ]DAT$fDAT$AD$HH f.kHcHVHT$HHHT$H>H&e( ke(?;HTe(kCHH HfDkkCHH fDAT$AD$HH H9@$[Iɚ;IH9'HD$PHHHHD$HHILHD$`pHILT$`HLT$`RHHLT$`zHP(Ht$PE1L$LT$PAD$@H\$0HD$8HQ(H D$(Ld$D$HL$HHD$ H0Q(HT$XH=%HD$Yb(HHLT$P HL$HHHD$PHHHtI*t1I,$'LHLT$PLT$P@LpfDHHT$H HT$HIfDHHT$HHT$HHfDLHrHHLL$LL$hLD$`L$L$kHILD$`LL$h=M I)LLD$`HD$hLD$`LT$hHHH|$HL$LD$hL$LT$`L$HHLT$`LD$hM5I(+LLT$`#fDH$L蓟HLnIHiJcHHF0H$ID$(H$ID$ H$ID$H$HIH I%Iu-H~1H5"J(HHVFH H$HH H$L$H$H$HD$XvHw"=a( #a(:HD$HE11H`(E1E111E1E1E1H/"`( `( ;HD$HH`(1E1E111E1E1IH!`( `(:E11E1Hz`(E111E1HD$HE1H!n`(# T`(;1H;`(1E1E1E1HD$HHk!1`(' `(\<11E1H_(H=L(HF8(H5O8(H6H={L(FH"H!_(# _(;H_(8IH _(# _(;Hh_(1E1E11E1HD$HMD$MIL$IHI,$uLHL$`LD$H LD$HHL$`LHHLD$`HL$HWFLD$`HHL$HI(fLHL$HHL$HOM_H5K(HHHVyHH$H51L(HHVUHH$ HH5 K(HHV-HH$ HHR^(* 8^(<H!^(E1MHMIXIHI(HCIؽ1ҿH"](' ](<<111H](E1NHL$HLAM HYIHHHD$`HHH>HCH\$H1ҿHL$XLD$PHLD$PHL$X_ H{A](' '](g<E111H ](E1HJ](# \(;IH\(1E1E11E1HD$HfH5?(H=Z(1_HHH踞H+uHzH\(( \(<Hh\(Hr\(' X\(<HA\(IL$HMD$HII,$uLLD$`HL$HHL$HLD$`HLHHL$`LD$HFCHL$`HLD$HHMHHHkH^HL$P,HHL$P3 H[(' [(<E1E11Ht[(1E1H=c&LL$hHL$`HL$`LL$hNHO[(' 5[(7<1E1E1H[(11LT$`HD$HlHB[(' Z(5<1E1E1HZ(11HD$HbH=fG(1IH=RG(H2(H52(NIHZ(' |Z(2<HeZ([IGHbZ(' HZ(0<E1H.Z(M$LL$`I H`&Z('  Z(:<1E11HY(1E1H=F(XIH=yF(H1(H51(uIjH=WF("HbH=CF(H1(H51(?H@H=q$LL$pHL$hLD$`LD$`HL$hLL$p1HLHL$hLD$`jHILD$`HL$hGHXY(' Y(<E111HX(H|$H1HHL$` HHHL$`HX(# X(<HX(oHKHLCHIH+6HLLHL$`LD$H?HL$`HLD$HHLHHHmH`Hn4X(# X(;1HX(HH8W(# W(;1HW(HW(# W(;HW(7H=HD(IFH=4D(H/(H5/(0I$HHT$HHT$HIH;%MmHHH|$HLT$hLD$` HCH\$H1ҿLD$`LT$hiLLL$`HCI1ҿLL$`LHH;z%IH59(H=S(1YHHD$HH\$HH菘HHD$PHHHyHfV( LV(;H5V(HHT$HsHT$HH~H_%V(  V(;1E1E1HU(H0U(! U(;1E1E1HU(1[HU( U(;11E1HU(E11(HU( xU(;HaU(1E1E111E11aHL@HH= LL$`LL$`>1tH$LFWHaH|$HxHT( T(~;HT(HT(( T(<HT("H$L-H5'LHHgT( MT(:H6T(5@T(,Ht:T('  T(V<111HT(H1LALL$`ILD$hH*S( S([;1E1E1HS(11E1HD$HGHH%H5H8nH=-eHmS( SS(:H-(HHLuLHHHLMLHHAHHHH; %HD$p HPHHT$p LpHIHHQHHHt$pH] H$IFH;%L H;%"IVB"HJ1ۨ HL$@uI^Ht$HkH-%H Ht$HQ;UP \"HHD$@H;H Uqp BR96 HHH|$pH$H H/uH$HHD$pWI.%H$H/HDŽ$BHL$ H@HHHt H9HPHuHHH@HHL$HHD$@tHHD$HHtHHD$@HtHILt$PIMHD$Hl$H\$XLl$PLH HHIHIHl$HD$HL-HD$HL$LLL,L0HLL"HLLHl$HuLl$PH\$XHL$ HH$HHHDŽ$tH+uHGHL$HHtHHD$HHHuH!HL$@HtHHD$HHHuHH\$8H5_(1HE:HHHSH\$HH HIHHQHHHD$ HHHL$(HtHHD$HHHuH|MtIEHPHIUuL^HD$0HtHHQHL$HHuH8H EfAMcIHHg-6(}6(bQE1HD$(H5(H|$`1CLCLCH|$pCH$CH 5(5(H=9%55(xfEfAFII ICfEf7EfAFII @GHHD$H|$H|$`H/4 *GHD$GHd$H D$H\$GHD$GHD$GHd$H D$H=%HH;%O9H|$p;L;5k%wLhHz4(w`4(6QE1HD$(H=4(HH9%HIfDfDHEJHHT$p`f.HEJ6HHHSH\$HH HHHQHHH%IHD$ HEJ0HHT$pf.JT5HDJTH~fIEJ0HHD$`f.IEJHH$IEJ<0HIEJ0H/IEJ+(}RE1HD$(H+(&fDH|$`ơ fDfDfDkfD'HH HL|$*(*(-SHl*(H|$`281HD$`"8H|$p8H$HD$p8H +*(E*(H=5(*(HDŽ$mH|$@H$HT$`Ht$p{jH$HT$`1Ht$puHHLt$1HL,IIHD$HHIu H|$vH+uHhML-I$HPHI$uL< H|$pI7H|$`HD$p6H$HD$`6HD$@HL$ LHT$8HDŽ$HEH%HD$ HL|$((((TH((HD$`Hi61HD$`Y6H|$pO6H$HD$p96H b((|((H=5_((HDŽ$NkH|$(HL$pH$Ht$`hHL$pH$1Ht$`HHLt$1HL.+HIHD$HHIu H|$譸H+uH蟸HH,HEHPHHUuHsH|$`;5H$HD$`%5H|$pHDŽ$5HD$(HL$LHT$ HD$pHDH%HD$(HD$ @HGL|$'(&(SH&(HD$`,DH L|$&(&(+SH&(+HL|$&(&(SHp&(HL|$u&([&(SHD&(HD$`H L|$D&(*&(SH &(mHML|$&(%(!SH%(lH!L|$%(%(7SH%(@H\$pH$Ld$`蕺HHHLBHHD$`HDŽ$HD$pu%(HN%(T%(.THD$(HL$LHT$ E1E1HBHD$("H[!%(%(TH$(H7$($(OSH$(HD$@HL$ LHT$8IE1E1HAH$Ld$`Ll$p艹HHLL@HHD$pHD$`HDŽ$i$(HB$(H$(dSgH^HHHHGLGHItIHu)M~)H5(LHV豹HtpH$IM_H$L$H5/ (LIHVoHH$uHEHF H$HEH$]H$LhH5'HLxHR#(F8#(PPH#(H`&#( #("TH"(H9"("(TH"(H"("(XSH"(H"("(SSH"(H=H|"(|b"(RQE1E1HD$(H<"(GH;ص%蠳IMuvHa'"(w "(*QE1HD$(H!(H.!(w!((QE1E1HD$(H!(誺H)H%H|$`I9L;5q%AA L;5%A X|I>H=(IH=(Hy'H5'I~IVHH$uInHHEI.uLHEI1ҽANH  ( (RH (DH ( (&TH (EHtV1nH (m (QHV (HL$8E1HHD$HHHFH9H%H5;HD$@H:HD$@H= (xIH= (H'H5'IH(u(QH\$0HD$(H(H=Ht$(Ht$(HDŽ$H~(d(QE1HD$(HA(LLxIHu;(!(oQE1HD$(H( %H8((mQE1HD$(H(H=e (H'H5'aI\H((yQE1HD$(Hl(wHv(\(tQE1HD$(H9(DH}C()(rQE1HD$(H(HJ(w(4QE1HD$(H(ɶIH (|(MQE1HD$(H(艶IH(v(}QHD$(HV(aL2HV(<({QHD$(H('H= (IH_H}HGHHH|$`HD$`H/uŭH$H|$`H HD$pH+EH虭8H=%HH8貭H((QE1Hm(H%HH8vHe(K(QE1E1H.(9LH^HH$H5Q(HxH9((QE1E1H(H ((\SH(H=E(HH(}(*RE1E1H`(kHj(P(SRE1E1HD$(H*(5Hn4(u(PH\$0E1HD$(H(hHuHT%H5H8转HD$`H (u(QH\$0E1HD$(H(H(q(PH\$0E1E1H_(HD$(aH`(uF(PH\$0E1HD$(H()HGHHD$p HWHHH$H$H/uЪHt$pHt;H$L&IH=Ht$=Ht$H$H(u(PH\$0HD$(H^(iHh(uN(PH\$0E1HD$(H&(11LMHHD$`R6HC (u(QH\$0HD$(H(H=n(H'H5'jH8H((ZRE1E1HD$(Hr(}H|(b(\RE1E1HD$(H<(GH5(I}ALHh.({(CQE1E1HD$(H((HD$հPH((iRE1HD$(H(H=F(IH=2(H;'H5D'.IHHʪ%H5kH83HM(3(RH(HL$HHD$HHHIE1E1H=Ht$HpHt$HOuLfH$LH(|(OQE1HD$(Hs(~Hs(}Y(_QE1HD$(H6(A,IHIE1E1H%HH8HD$`H<((SH(HL$HHD$HHHt\E1E1HD$(H=%HH8貦H((SE1E1HD$(Ha(lHE1E1>HD$(PHO(5(QRE1E1HD$(H(HS((RIE1E1H(HD$(D'H5'H=(1HH$t]HVH$H/uzHHDŽ$(s(SE1E1HV(HD$(XHW(=(SE1E1HD$(H("H[!((~SIE1E1H(H+((SH(HT%HH8ɤHD$`H((RH~(]H%HH8臤Hv(\(RIE1E1H<(GHF(u,(PH\$0HD$(H(H;%H|$0fHHH$uiH((SE1E1HD$(H(H((RE1HD$(Hr(}H;=v%H;=%H;=/%茨x!H$kH|$01ZH2H^$( (SE1E1HD$(H(H(((RE1E1HD$(H(H((RE1HD$(H{(H(k(RIE1E1HK(HD$(MHL(}2(]QE1E1HD$(H (9HF ((HTE1E1HD$(H(LwMtOHGIHH$H$H/u臡H$1AHG=E1!Hb(H(RIE1E1H((HD$(*Hc)((kRE1HD$(H(H&((RIE1E1H(HD$(H=M'HlH=9'H"'H5+'5HJfffff.AWAVAUATUHSHHH*%HH|$XHDŽ$HDŽ$HDŽ$H$HFH H HF0HD$`Lu(H] L}Hd( H(hE111AHLHHD$PHHHp H( H(hE111AHHHHHH: H( L(hE111AHLAHIHH Ht$PE;Fu A;D$ HT(H5'H9pH'HHL5'MIFH5(LHHHH2I. H (H ['H9HH:'HvHL&'MICH5'LL\$hHH{L\$hIMI+} IEH;%AE1E1H;%H;%L\$h蒡HIL\$h#MtLXHD$PIHKDHEKlIELMLD$h^L=Ϡ%H LD$hQA;P 1LLD$hLAI(x LD$hWP A=H @9M<I(|Im HCH;f%!H;%L$;H;% #HSB #HzE1 H|$h膠L=%x WA;P "LLHD$hIXH Aqp B9M%MHI. MCH) L;=%L;=% IHPHI H8 (H5'H9ptHp'HPHH\'HHCH5(HHHIMH+= H (H 'H9HH'HHL-'MIEH5'LHHIMIm I@H;o%A1E1H;՞%'H;% LD$h耞HILD$hMtLhHEHIlI$OdI@HHFL\$pLD$hKL=%p LD$hL\$pVA;P 1LL\$pLLD$hH H LD$hL\$pQP A=H @9 HFI+ I( IFH;F%H;%H$H;|%O!IVBDLk"@$@$@$DHD$PHHLL$L\$hH$H$C8HIL\$hMI+L臋fHHLLD$hL$H$L$7HHLD$huM{ImpLLD$hLD$hY@HD$PHHLL$LD$pL\$hL$H$f7HIL\$hLD$p`M"I(LL\$h蠊L\$hfDH$L 7IzH$H6I:H$H6I`LnIVHzJcHHF0H$HE(H$HE H$HEH$H-IIIIu-M~1H5^'HHV肎H&H$IMH$L$H$L$HD$`Hy' _'1E1E1E1H?'E11E1E1Hx>' $'1E1E1E1H'E11E1VH@' '1E1E1E1H'E1E11E11LD$X0HLD$XH' '3E1E1E1Hw'H' g'O3E1E11HH'E1HO' 5'Q3E11E1H'p IMhMIXIEHI(uL̇HCIA1ҿH=g'HP'H5Y'cIrH' 'L3E1E1E1Hq'E1H='ݭI+H='H'H5'IH5L'H='1nHH H*8H+uHHE ' '3H'E1E1E11E1E1#H ' '3I1E1H'E1E1L\$`LD$XHLD$XL\$` H|' b'|3E11E1HC'H=7L\$pLD$h踉LD$hL\$p1LLL\$pLD$h=HHLD$hL\$puH+' 'q3E11H'MFMINIHI.uLHL$pLD$huLD$hHL$pLHHLD$pHL$hLD$pIHL$hI(FLHL$h/HL$h/MH5'HIHVHH$H5='HHVHH$ IH5E'HHV虉HH$ ILCMHKIHH+uHHL$pLD$hnLD$hHL$pLHLLD$pHL$hLD$pIHL$hI(LHL$h(HL$hH5W'H='1iHHc H%5H+uHH@' '83H'H' '&3H'HE1E1E1E1蕋IHֳ' 'J3Hk'E1E1E1E1E1H='˩HH='H'H5'HH5*'H=s'1\HH H4H+uHڂH3' '4H'H ' '3H'藊LD$hIHӲ' '3E1E1E1H_'E1H='˨I9H='H'H5'IHj0' '3E1E1E1H'E1MHH*' '3H'E1E11E1E1 H=U' ILCMHKIHH+uHLD$pHL$hWHL$hLD$pLHLLD$pHL$hLD$pIHL$hI(NLHL$hHL$h7L\$X荅HL\$X HH' '3E1E1E1H'.H=ȺL\$hNL\$hMCMMsIII+IFMA1ҿAH|' b'3E1E1E1HB'1LLL\$hkHIL\$hjH^$'  '3E1E1H'GH='H'H5'荦IH=o':IH' '2H'膇HHǯ' s' 3E1E1HV'H`' F'2E1E1E1H&'E1}M]MdMuIIImIFM1AAH='OI1LLLD$hHILD$h蚆L\$hI}H=6'H_'H5h'2I%Hz' `'2E1E1E1H@'E1H=1LD$h跁LD$hsH^$'  '2E1E1E1H'DLL\$hM}AIF1AL\$hLLD$hA}IFM1ҿALD$hH̭' x'2E1H^'Hh' N'2E1E1E1H.'Hr8' '2E1E1E1H'1V[H5'H=C'1,HIH-ImsH' '2H'H5'H='1HI$H-ImHq' W'm2H@'aH5d'H='1vHIH2-ImHP' 'M2H'"D$xOHT$pL$xH' '52H'Hܫ' '2E1E1Hk'HT$pnH[' A'+2H*'KD$hHT$hHN' '!2H'H=sH ' '1H'5''LH蟙I~HuH|%H5H8S{E1xHj' P'k3E1E1H3'H1LLD$hHH`&'  '3H'k~HuHW|%H5H8zE1AE1H=4}/E1kfuLH藘IRH$HI5LFyH`' F'i2H/'E1E1E11E1{H=(HJ' '1H'9H#' 'c3E1E1H' AE1H{%H5H8oyL\$`LD$XH' e'3E1E1HH'uHLHI#H$LwIHX' '4H'}HYz%H5H8xL\$XH' '3E1E1H'H1LALD$hIL\$pLfwjH' f'I2HO'L2wHL' 2'2H'A1;Hpy%H5H8wLD$X=LH ' '2E1E1H' H' '43H'H$LH5_'LHHi' O'1H8'zHHx%H57H8vnH=y!H:' '3E1E1H'#H1LAL\$hIH=0yq0H$HIffff.AWAVAUATUSHH(HF HFHHtHHcH@E1HCH5'HHH IM @zHH HgwHHH5T'HH{HmH'H 'H9HH'HHH-'HHEH5'HHHIMbHmCH5|'LH{I,$.IFL%#'HH# wL=Ow%H QA;P K HLLIwH QP A=H @9M! I.H+ID$H5'LHH HHI,$HCH;u%jHkH]LsHEIH+H'HLRHmIuHrMI.MID$H5'LHHHHI,$%HCH;u%HkHLcHEI$H+ID$H;kv%Hl$pH;)w%\IT$BIE1 HZuMt$vL=u%H QA;P =HLIuH Aqp BR9MHmuHqMEI,$LIFL`pMWI|$KLtHHH=Lu%HHxHmHuH7qH&HLAT$H+HuHqHI.uLpHH([]A\A]A^A_fHpMHpLpt@$$fHp2LpLxpJLhpLXpHHpOL8pH(p+t@$ZfDnAMcIpImLHHIHHDnFII Dnf.DnFII IDHPtIH'r'HE1H'MtI.uL`oHtH+uHMoHtHmuH9oMtI,$uL%oH .'H'H=5+'1$!HQ'r'W1H'\Ht$LFIH;r%H-`'H,$H;s%HHSBbE1 LbuLsvrL=q%H QA;P HLAILrH Aqp BR9|MI:f. r@$HHuIH;r%H;r%ZHSBKE1 HjuLcqL=$q%H QA;P 1LIqH Aqp BR9|MIfKq@$11HIHPHYp%H5:H81tI.H]#'s 'H'LlHLI H's'TH'"1nqHyHҜH=kY'Ym'HHS'^1?HV's<'mH%'Hi/'s'H'HB'r'FE1H'LH=x'H'H5'tHJH=V'!H6H's'RE1Ht'jsI+HLn'sT'E11H8'HoI sI)H_%'r ':E1H'H5'r'DE11H'=LHLqHI!H'r'aE11H'H=sm@nHuHl%H5hH80k9rHHz@'s&'|L1H 'rHOsnHE1#Ht$L(IH=Dm%nHfE11H=mH5'1HImHtSE1/HHZIH=7lϨfDtHHnIHwk%H5H8iH\k%H5H8iHAk%H5H8iTDAWHAVAUATIUSHHx10IM0HA'H5Rl%IEHGl%HD$@HD$HHD$PHD$`IEPHHD$Hl%IEXHD$HMHHn'HD$XHCHHHHH$HHj%H H5L MHOH81LDHoH'z'__zHj'H ϗH=gImuL8g1HC8HD$`HC0HD$XLCHC(H{ HD$PLD$@MH|$HL5qk%L9wHGHPHHwHHt$ H|$`HD$XL\$PHHD$(MH;=j%H;=4j%‰D$4L;=sj%IWL;_j%pIIGHTHt$ AE0IuHaH|$ ICH;h%L\$fDH\$ H|$ %rjHHD$ H;=i%2L\$kD$4L\$*jHL\$ H'{'__{Hp'D$4HH5M'LL\$HL\$HHc Hg%H9BMHJH@HZHHH*WH'HHL\$8HL$9HL$HD$L\$8H)H|$OH+I+HD$H@H;hg%HD$HI}XH/rHD$H;Oh%IEX(HH Ic}0I](HoeIcU0HIE8HIU@IHII1+DHGHHtzHIE8IcH,I9W~zI\AHHL9sHxH; HCHHH H HkuHHB H;HGHHuHHT$ecHT$oIHHI^H^'HD$(H9L5h%L9pL9sE=HD$(x  { SHD$(H@H;CHJ'HD$(H9L9pL9sd\HD$(x ! {  HD$(H@H;C+H='H;=nf% HD$(H@f Ht$(jIM" H='LHI I/uLbLCI.uLbHz$' 'aE111H'E1G@Ht$(H;5e%uH;e%uH|$(H'bHI) H;e%H;e%I/uL^amHW'HI}PH/AM0H:'1I}@Iu8IEPHD$ ~HHH9IE D$4H5'H|$AEhpaHIH;d%H;cd%HIHPHID$4A]lHt$1HHD$HHHuHo`LHx[]A\A]A^A_HL$(H; @d%uH;/d%uH|$(H`HIlH;(d%H;c%I/uL_7NH+'HI}PH/LAU0H'I}@Iu8IEPHD$ @HcʃHH΃uuL;5ac%LdcHHa'G'a11E1H)'@Mu L4dHHIEoZLHH|$ HGH5b%HHHP'BL;-)K%uH;5K%uLGHHcBH;K%H;J%e,HmuHF-B^f.LFHFLFHxFLhFpHXFKFfDHHD$3FHD$fHHD$FHD$HH|$H5'HGHH&SHHRHCH;J%-H;~H%x.H@h1HHYH@HYHHCYH+a'H|$H5u'HGHHYHHXHCH;`J%*-H;G%.H@hHHNH@HMIMMH+&LH{EHHwSHm&I,$&H;H%H;XH%%HHPHH&pH|$ H5'HGHH6HHX6HCH;nI%H; G%!H@h1HH-0H@H 0IM6H+H|$H5'HGHH6HH]5HCH;H%5H;F%!H@h1HH/H@H/HH=8H+yHL DHH7I,$BHm'H;pG%@H;F%@@H++6H|$0HWHBpH.H@H.H5)'HH.HJHHD$87H+H|$ H5'HGHH8HHn7HEH;G% H;EE%_ H@h1HHz.H@Hm.HH.HmgH|$8HW HGHHH9<H9<HHWHHHGH+4IGH5I'LHH~<HH<5HqD%H9E[-H]HN-LeHI$HmHT$8HLH+HuHAH4I,$HEH5'HHH3HH!/HmH|$ H5*'HGHH.HHE.HEH;F%H;C%H@h1HH,H@H,IM,/Hm"HKC%H9C.HkH.LcHEI$H+1ҿAID$H;D%$H;IE%K6DHIJ.HtHhH'IHKDM|ID$HHX1LL$ DHzC%P LL$;P 11LLL$LHD$(CH LL$QP =H @9H|$(0I)I,$H|$H5'HGHH/IMD/HŮ'H 'H9H /H'H.HL 'M.IAH5'LLL$HHs.LL$IM-I)HiA%I9D$:*I\$H,*Il$HHEI,$LHHH+HD$uHv>I.,H|$,Hm5HT$HHD$@HHHH5a'L9 HB%I9]H9^= 5 A} :~ JIEH;FH5'L9 I9]H9^~vA} ?~ ~JIEH;F{#;HIIHf'H='LL$HIAHK'HHy' EHILL$QVCHHLL$QE1LHLHBHmILL$u HA} A~ GIEH;F!H'H 5'H9HGH'HGHL='MEIGH5\'LHHxEHH1EI/H|$@DH|$x8HIEH|$@1H;HHPDI/H=%H9CCHHIHmM}CH+]L;5F>%L;5=%IHPHIH=%H =%HH=%H9s=%HL$xH5'L9H>%I9EH9F%A} ?8~ BIEH;FH5'LBBH|$pfXH'H5 'H9p[H'H[HH׀'Ha[HCH5'HHH6[IM`ZH+ H|$@YHT$@H5(=%H9rYHBH+YHHH+;%I9AXLHLL$HLL$HIHmn MDYI)K H|$pL8HIpWI. H'H'H9P5H'Hz6HH}'H66HCH5i'HHH 6IM5H+& IFH;C:%,5E11H;;%H;m<% Z;HI4HtHXHD$(IHKDIEMlIFHH40;H:%P ;P T41LLH;H QP =p4@9H3I/I.N HL$(HHD$HHHH! Ht$ H5HHD$(]3Hm H|$8 5HI3H|$(H5Q'LLL$HHGHH2LL$HI)uL6HD$(HHHL$0HtHHD$HHHHuH5HL$8HtHHD$0HHHuH5HL$(HtHHD$0HHHuH5HL$HtHHD$(HHHuHt5HL$pHtHHD$HHHuHN5HT$@HtHHD$HHHuH(5HL$XHtHHD$HHHuH5HT$hHtHHD$HHHuH4HL$`HtHHD$HHHuH4HT$PHtHHD$HHHuH4HL$xHtHHD$HHHuHj4MtImuLV4H|$ HHD$HHHu84HT$HHD$HHHuH4H{HL$HH; 7%u:H;57%u)H|$Hb4HI2H;7%@H;T7%@H@I,$uL32HD$xL;-a7%uH;5P7%uL3HH 2H;N7%@H;6%@$@H+uH31H|$PL6HD$PHIL;-6%uH;56%uL13HI.H;6%@H;#6%@3@I,$uLd2.rZ0HHP9Hw'H='HHCHa'HH-?'9HI8!8HI8E1LHHHa7I.IuL1M[8H+] ID$H'LHHH7HD$H|$7Ht$HH' I,$1 HD$H4%HH9P|6H\$Ht$HIM5H+ ID$H;3%3I|$H)3ID$HD$pID$ HD$@ID$(HD$XHD$pHHD$@HHD$XHI,$ HD$PHD$`HD$hDH;4%ZH5LH`''Y?Ho'HD$E1E1E1E1E1HD$xHD$PHD$HD$`HD$hHD$XHD$@HD$pHD$HD$(HD$8@HtH+uHLL$H/LL$HMtI,$uLLL$H/LL$HMtI/uLLL$H/LL$HMtI)uL}/MtI.uLj/H s''H=v5p'1i8@H8/'H(/L/H/zH.H.H.H.L.FH.H.]H.HAm.1ҿLX.LH.sL8.H(.H.L.lID$@H'HHLH$L$H$*HHD$(KH脪I/L-DH-Hp-L`-HP-H@-$>+HH+;HT$0Hl$0HHEH8H,IF@nHD$(HHLH$L$H$HHHHy@L,LL$0LL$@$HL$xHt$H^HL$xJfDL;-10%uH;5 0%uL,HH-H;0%@H;/%@F@H+uH+-2)HI9H'H=1'HID$H'HH'B3HI91HH9E1LHLH0HmIuHHD$:+LL$M9I,$A IAHf'LLL$HHHR9LL$HD$hH|$h=HL$hHH" I) HD$hH \-%HH9H5=Ld$hHt$LIM<I,$5 IAH;:-%;IyH_;IAHD$@IA HD$pHD$@HHD$pHI) HD$PHD$`HD$XHD$L;--%uH;5-%uLi*HH:0H;-%H;\-%HmuH)0lH'H q'H9HN:Hq'H*:HL=q'M9IGH5'LHH9IMn9I/ H'HOq'H9P9H.q'H8HL%q'M8ID$H5e'LL$HH]8L$HH8I,$ H|$X7H|$XH5'L$HGHH5L$IM5H|$@5Ht$@LL$(HHL$$4I,$o HCH;*%/DŽ$E1E1H;,%qH;,% L$+HIL$;MtL`HT$XIHc$KlLHL$HIT1˙HIL$L;I.U H+ L$!+HHL$w/LxHD$L$HHC -HIL$ /HT$xH5?'HL$-L$HT$xH5Ë'LL$-L$LLHL$ݘHHD$xL$.I)H+I/ Ht$xH;5]*%H;5+*% H;5)% ʄOH|$xm+=H^V$'7 'BE1E1E1H'E1%fL%)@$fHCL I$HCH(HEHEHHHEL8IHX%H|$F%I,$L3%fDH %L%HD$PHD$`HD$hH5I~'H|$HZAZH5v'H=.'1HH:HHm:HT'1'WAH'E1E1E1E1HD$x@Hѓ'H l'H9H7=Hk'H:HL5k'M9IFH5{'LHH9HHY9I.V H|$@8Hs&%H9C#BHt$@H IMAH+ IAH5+u'LH@pH6H@H6LL$HLL$HHHbCI) H|$XHBHt$XH#HIAH+ H%#H5w'H|$HJD3@Hu'Hj'H9P?Hj'H!>HL5j'M=IFH5w'LHH=HHt=I.* H53t'1HٔHI:@H+ I.LQ"@LcI$$f.HkHEfH]HL}IFIMHVH9tHt HEM DF DD@9\A 2 I}0MUHA@IDA  LN0HHA@LD΃C  7A 9 H!L1HJ$yIMHVH9tHt HAEM DF DD@9A  I}0MUHA@IDA a LN0HHA@LD΃  7A 9HGL1H#@#HL$HHVHIH9tHtHfDHL$HDF DI D@D9kA  L\$HLMSHH0A@IDA  LN0HHA@LD΃  7A 9H9L1H"@fIMHVH9tHt HEM DF DD@9A  I}0MUHA@IDA  LN0HHA@LD΃ R 7A 9HL1H!@H;"%H#$HO؍'')?H'3fL;5"%rL#cHN'6v'AE1E1E1HV'HD$xH((HL:HHLLLL$LL$fLH|$hLL$LL$LKLxH|$`fLX LLL$CLL$fL([LL$L$LBLL$L$:HHsL9LL$L$tHL$pL$ Hy %HHD$xL@@HC@\HD$XHHHL$L$H$H$]HIL$^4LL$詘HmL$iHL$T1HHD$0HK' '>Hy'} @Ho'HHLH$L$H$HHD$(2HsHmhH[LL$L$HD$(HHLH$L$H$8HH2HH+HH;-m%HL;%Q%LL;%5%LH 0IMHVH9tHt HhEM DF DD@9=A I}0MUHA@IDA LN0HHA@LD΃7A 9HL1Hb@IMHVH9tHt HEM DF DD@9A I}0MUHA@IDA LN0HHA@LD΃#&7A 9HL1HLHt$xH;5%H;5% H;5[% :fHHL8HCH(HEHCL`I$fDLHLL$HLL$H{fHLL$HLL$HfHLL$HLL$HfLLxLhCLXHHHkHEfLc I$fHyG?'J%'*CE1E1H'FH;%H>H0G'4܅'AE1E1E1H'HD$xHF'4'AE1E1E1H'HD$x}HD$XHHHL$L$H$H$HIL$@HrF8'4'AE1E1HD$xH'HHmHLL$HLL$HH;-%HH;%HLNHI}HLNHI}HLNHqH|$HHHILNHI}H7A 7A s7A D7A LNHcI}H@LNHI}H7A k7A 7A 7A 7A #7A 7A 7A LLLHD$Ht$8HIHH%HRH5jsH81HDW'='z?H&'H;Hh@;Iw;Hg;I LsI,HnJcHHC8H$HC0H$HC(H$HC H$HCH$HIIH%JcHHCs'Y'?E1E1E1H9'E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$E1HD$HD$xHD$HC'ȁ'?E1E1IH'E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$E1HD$HD$xHD$gH|$ 8HHyB?'%'?E1E1E1H'HIB''?1H܀'IܺE1CIܺE11'HAI''?E11Hw'9HA'!g'?E1E1IHG'E1HD$PHD$`HD$hHD$XHD$@HD$pHRA'!'?H'HD$HD$E1E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pHD$LL$IH@'!m'?E1HS'H=k'6IH=k'HX'H5X'6IHa@''! '?E1E1E1H~'HD$PHD$`HD$hHD$XHD$@HD$pHD$E1HD$HD$xHD$H|$ILL$HLL$'H?x~'^~'?E1E1E1H>~'HD$HD$xHD$PHD$`HD$hHD$HD$XHD$@HD$pHD$HD$(H=HLL$]LL$R1LLLL$HHD$(LL$(H>}'}'Q?E1E1E1Hf}'HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$8Hb>(}'}'L?H|'E1E1E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$8H|$ HH(H=|'w|'?E1E1HZ|'H|$KHLH=R|'8|'N?H!|'E1E1E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$8HD$E1HD$HD$xHD$H<{'{'?H{'E1E1E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(H5]'H=x'1x~HH#H4HmuH HN<{'z'h?Hz'H'<z'z'V?Hz'E1E1E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$8oH;uz'[z'S?HDz'H;Nz'4z'?Hz'Ha;'z' z'|?E1E1E1Hy'E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(H|$ HH]HLeHI$HmuH[ HT$HLL`HD$HHHH`H( SH|:By' (y'i>E1E1E1Hy'HD$PE1HD$`HD$hHD$XHD$@HD$pHD$HD$(HD$8HD$0HD$xHyH9x' x'g>E1E1E1Hex'HD$PE1HD$`HD$hHD$XHD$@HD$pHD$HD$(HD$8HD$0HD$x=H=d'n/I颽H=d'HxQ'H5Q'/I逽H 9Lt$ w' w'Z>E1E1Hw'IE1HD$PHD$`E1HD$hHD$XHD$@HD$pHD$HD$(HD$8HD$0HD$xKH]HqLeHI$HmuHLHLO^HD$ HLHHHE1E1Hv'E1E1HD$PHD$`E1HD$hHD$XHD$@HD$pHD$HD$(HD$8HD$0HD$x1$HPHe7+v' v'I>Hu'Lt$ E1E1E1HD$PHD$`HD$hHD$XE1HD$@HD$pHD$HD$(HD$8HD$0HD$xH=*b',HuH=b'HO'H5(O'-HSH54X'H=r'1xHHI(HBHmuHH\6Lt$ u' u'7>E1E1Ht'E1E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pHD$HD$(HD$8HD$0s HzH|$8H7H5dt'Jt'?E1E1E1H*t'HD$E1E1HD$xHD$PHD$`HD$hHD$HD$XHD$@HD$pHD$HD$(LHt$Ht$&H4s' s'>Hs'Lt$ LE1E1E1E11E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pHD$HD$(HD$8HD$0LHt$ Ht$H64r'"r' @Hr'E1E1E1E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pHD$H|$(LL$H H3mr'JSr'(CE1E1H6r'tHz3@r'I&r'CE1E1Hl$(Hr'E1?wH"H73q'Hq' CE1Hq'H=<HH2q'Hq'CE1E1Hq'I^HI$InHHEI.uLIHEI1ҽA1LLn HHH/H=]'HI'H5I'(HHD2 q'Hp'BE1E1E1Hp' IH2p'Hp'BE1E1E1Hp'H=7]'(H}H1p'op'?HXp'LHt$HVHt$HH1Hp'>.p'^BE1E1E1Hp'E1IH|$HH$H$H-1o'/o'/AHo'7H1o'o'>Ho'H|$ HɶH0o' yo'>E1E1E1HYo'HD$E1HD$BHH0Io' /o'>Ho'E1E1E1HD$PHD$`HD$hHD$XHD$@HD$pHD$(HD$8HD$0H=b['-&H>H=N['HH'H5 H'J&HH/n' xn'x>E1E1E1HXn'HE1HD$PHD$`E1HD$hHD$XHD$@HD$pHD$(HD$8HD$0HD$xH9/m'm'?Hm'H|$HYH5P'H= k'1pHHVH讯HmuHoH.m'tm'?H]m'H.gm'Mm'>H6m'LHt$4Ht$\H`.&m'% m'@Hl'%HHxH%.l'$l'A@E1E1E1Hl'E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pH;k%LHHI,$HEHHHHD$pyHHHD$@\HHHD$XHӾHf/HmHHD$PHD$`HD$hH%H/H58WE1E1E1H81H,k'?zk'hBE1H`k'H5t_'HHVHH$mIMH5S'HHVHtH$IM~`H5`'HHVHtH$IM~7H5Q'HHVHt H$IIQMH$L$L$HD$HH$HD$xH$HD$0H$HD$鍮H+gj'$Mj'8@E1E1E1H-j'HD$PE1HD$`HD$hHD$XHD$@HD$pHD$xHhHHtHXHEHHHD$@HHHPHT$HHQHmIFH{9H$H8GtH!H*zi'#`i'@HIi'E1E1E1HD$PHD$`E1HD$hHD$XHD$@HD$pHD$xHD$FH3*h'0h'AAHh'=LH$H$XH)h'3h'iAHh'H)h'#qh'@E1E1E1HQh'E1E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pHD$HG) h'#g'@Hg'E1E1E1HD$PHD$`HD$hHD$XHD$@E1HD$pHD$HD$xH(g'/xg'5AHag'HH$\H$aSLcMLsI$IH+IFLDŽ$1ҿAH8(f'4f'AE1E1E1Hf'HD$xH'f'4f'AE1E1HD$xHf'H'f'4uf'AE1E1E1HUf'XH'_f'Ef'#?H.f'mIHe'+f'@f'}BE1E1E1He'E1,HHt$HHt$HGH'e'6e'AHD$xE1E1He'E1E1LcML{I$IH+HLLLII$LHHI$L)ٻH}&Ce'6)e'AHe'E1E1E1E1HD$xH$H)H5PE1E1E1E1H81=H&d'6d'AE1HD$xHd'H%d'6d'AE1E1E1Hod'HD$x\H逺H%cd'6Id'AE1E1E1H)d'HD$x^Hd%*d'6d'AHc'H=%d'+c' AE1E1Hc'HXHH.LHHIHHD$HHHMHT$LHLL$JH+HD$PLL$HLL$MLL$H$H8 H$Kc'*1c'AE1E1HD$PH c'EHO$c'4b'AE1E1E1Hb'HD$xH$b'*b'@E1E1HD$PHb'HD$`HH$H$-H=O'IUH=N'HG;'H5P;'I3HHt$6Ht$hHHt$Ht$H>#b'a'?Ha'H|$HͬH#a'4a'~AE1E1E1Ha'HD$xH|$XyL$IHb$H%H5LLL$HE1E1E1H81H"Ka'41a'AHD$xLL$HH a'HP"a' `'>H`'H)"`'&`'@H`'LL$HD$hH!`'&`'@E1E1H}`'H!`'m`'&?HV`'H!``' F`'>H/`'HHt$-Ht$aH$H<)H5KLL$HE1E1E1E1H81KH$!_'4_'}AHD$xLL$HH_'H _'4_'zAE1E1E1H{_'HD$xhL$HH g_'4M_'xAE1E1E1H-_'HD$xbH=K'L$L$IH=K'H 8'H58'L$L$IH ^'4^'uAE1E1HD$xH^'ֿIFH^'4{^'sAE1E1E1H[^'HD$x鐿H=J'IH=J'Hk7'H5t7'IHXHxLL$LL$HB^'(]'@E1E1E1H]'HD$xHD$PHD$`HD$XHD$@HD$pHD$;H;$ILLL$0HHLL$I)HCHHHHD$@nHHHD$pHվH -H+H[HA]'(\'@E1E1HD$PH\'HD$`E1HD$XHD$@HD$pHD$HD$x鮽HXHHL`HI$HHD$HHHKHT$HLCH+IHLL$LL$wH$LL$H8LL$zHH\'&['@E1E1HD$PH['HD$`HD$`LL$HLL$HHH['['!?E1E1E1Ho['H|$`HHg['M['?E1E1E1H-[' HwHd*['4['AE1E1HD$xHZ'H.Z'4Z'AHD$xE1E1HZ'HZ'*Z'AE1HZ'HLL$tHZ'(hZ'@E1E1HD$PHBZ'HD$`|IyHIAHH@HL$@HD$pHHH$HH5EH81MH&Y'FY'BHY'E1E1E1E1E1鯺HY'FY'BE1E1E1H|Y'麺rHaHyY'F_Y'BE1E1E1H?Y'}H=E'IHqHb(Y'1Y'SAE1E1E1HX'E1HD$x H&L|$pLd$@X'(X'@E1HX'E1E1HD$xHD$PHD$`HD$XHD$@HD$pHD$lH+uHAGuH`HLd$@JX'(0X'@E1E1HX'E1E1HD$xHD$PHD$`HD$XHD$@HD$pHD$1cL`H$H57(LL$H81LL$HW'~W'd?HgW'kHqW'WW'?H@W'E11E1HD$PHD$`HD$hHD$XHD$@HD$pH$H5'H81PHL$H5!H8LL$RHV'HV'BE1E1E1HV'铷HV'4V'AE1E1HD$xHZV']LHHOV'5V'>HV'"H=B'H.'H5.' IH@V'U'6?HU'HLL$XL|$@H\$pU'$E1HU'U'e@E1E1HD$xHD$PHD$`HD$hHD$XHD$@HD$pZHmuH.4uHMHvL|$@H\$p2U'$U'm@E1HT'E1E1E1HD$xHD$PHD$`HD$hHD$XHD$@HD$p鹵ZHD$@1JLLHT'$T'[@HhT'I|$HdID$HHL$pHHH@HL$@HD$X5Hy?T'9%T'BE1E1E1HT'CHFH<T'9S'BE1E1E1HS'H=\9'7 IH$LH5I&LH&HS' {S'=HdS'5nS'雜H*&HH0HVHtH$IH$L$HL$HH$HD$HL$xH$HL$0/HC S'HR'BE1E1E1HR'ҳH=_dHR' R'=HR'$HR'0}R'GAHfR'۾HIHL$L1AIFDŽ$1ҿL$%DŽ$E1H=7'H*'H5*' IH5b4'H=+O'1UHItIHԓI.uLHQ'<Q'ABE1E1E1H{Q'E1鶲HQ'<hQ'=BE1E1E1HHQ'醲HRQ'98Q'&BE1E1E1HQ'E1HYQ'Q'?HP'H2P'FP'BE1E1E1HP'E1龱HkHнLcHEI$H+uH}HT$@HLL7IHEHHHEHLL$HELL$H邽HZP'F@P'BE1E1E1H P'E1 H$HIH5;H81iHBP'FO'BHO'HO'DO'BE1E1HO'H$HH50;E1E1E1H81HO'DrO'BE1HXO'閰H;HLL$)LL$H$H5AH8 &H]#O'F O'BE1E1E1HN''IYHFMaHI$I)uLHHL6IHMHHHHLL$HxLL$HHLL$H HD$@xBLL$HHHH`N'DFN'BE1E1E1H&N'dHj0N'DN'BE1E1HM'7H|$@LL$H%LL$HHH$HdH5f9LL$HE1E1E1H81HM'DM'BLL$HHM'ŮHM'DwM'BE1E1E1HWM'E1WH|$@HHLL$HHXE1{H]#M'8 M'BE1E1E1HL'E1$H*L' L'3>HL'I¤H L'DL'BE1E1E1HL'E1齭H=#9'HOH=9'H%'H5!%' H-@AW1AVAUMATUSHHH|$xH$@HL$ HHt$HH$IHT$HL$HDŽ$HDŽ$HDŽ$HH*L'H !'H9HaHr!'H=HH=^!'HH$HGH5A'HHHH"H$H/bHDŽ$ HH$HL$HHHHH$H$H5:?'HIH$H$HNHH$H+7H$H/H$HDŽ$H/H$Ht$HDŽ$HD$`HHD$HHHH|$`HDŽ$H57'HGHH7HH$VH5 /'H9tHPH;S$-HxH$HH\$HH$HHQHHH$H;|$HDŽ$H;=$HHPHHHDŽ$BHD$(H HL8Mt L;=$(HPHuMHhHXtIHtHEHtHH\I'H5'H9pH'HHH=p'HH$HGH59'HHHH$H$H/_8H|$`HDŽ$H5g8'HGHHHHH$H9AHD$LaM}LqI$IH)7LL0H$I,$uLH$WI.7H$HD$H9GH$܆H$H$H/B7H$HDŽ$H$H/7H$MHDŽ$HDŽ$HD$ptI/uL HtHmuHHtH+uHH5+'H|$p`HH$AH;D$HH;M$B5HӃHPHHa5HDŽ$HF'H5'H9pH'HHH='HH$HGH571'HHHH$>H$H/&H$HD$HDŽ$H9GHt$HDH$H$HtH/A&H$HHDŽ$H$H/"&H5*'H$HDŽ$H9tHGH;$H$H$HH$H/ HH$H;|$HDŽ$H;=$[FÃHHPHHGHDŽ$/L;-$LHHD$(WHE'H 'H9H+H'H=HH='HH$5HGH5,'HHHH$|H$H/'HD'HY'HDŽ$H9XH,'H"HH'H4HCH5T6'HHHIMH+q'HD'H5'H9pϾH'HHH'HHCH55'HHHHH`H+_'H$I9D$HD$ʹHLUH$Hm'H$I,$&H$H55'HGHHpHHoH$H/&H$HD$HDŽ$H9GH躁f.H$H$Ht H/u@HDŽ$H+"&H$H$H/Q&H5bA'I}H$HDŽ$HDŽ$H9HD$h8 HXH HJH~+H;r 1fH;t HH9uH%B' H(hE111AHLHH$H$HH$H/%ImH$HDŽ$HD$Xs%HD$XH;$HDŽ$H|$XH5q.'HGH_HHHH$H5%'H9"HPH;$Hxux!@HT$HH$HHQHH$H$H;|$HDŽ$H;=$HHPHH-HDŽ$̹H|$XH5('HGHHCHH$?Ht$pHMHH$CH$H/-H$H;|$HDŽ$H;=$HHPHHHDŽ$HHt$(='H?'H'D$(H9XH'HUHH='HH$MHGH5=/'HHYHHXH$H/,D$(HDŽ$VHH$qHt$H9sHH}H$H$H/Q,H$HDŽ$ʽH+,H$H;|$H;=$zHHPHHHDŽ$HW>'H P'H9HH/'HHH'HHCH5,'HHHHH$CH++H$H5X''HGHHMHHSH$H/+H5="'H|$X1HDŽ$HH$Ht$H9sKHkH>LcHEI$H++H$HL$H$HmuHH$H/*H$HDŽ$I,$*H$H;|$H;=$HHPHHHDŽ$D$( ,\,fTHH$'Ht$hH^HHH$H/@*H;\$HDŽ$H;5$HHPHH)HD$HHH;T$*H;'H5'H9poH'HHHs'HHCH5&'HHHL~HH$C~H+HH~HD$HHHCHH$XH;'H5'H9pH'HHL%'MID$H5*'LHH)IMI,$H5-'H$L+kGI.H$H$HQ=HHD$PH$H/HDŽ$H+H$H/Ht$HHDŽ$HHD$HHHHD$H\$ H9H; $ # H|$XH;=?$*HGH56-'HHFHH$BH $H9HHL$PHXHCH@HHH$H$H/"(H$H xH+IuHMH$H/'IGH;$HDŽ$1>H;*$DHXhHBH{7HHLHHpLSI H5$H2$HHH\$JH|$`H5 'HGHHHH$HPH$HHRpHeHRHeHH$!H$H//'H$H5'HDŽ$HDŽ$H9HD$p HH@H;$H{ H$1H=$HHD$H9$H$HHPHH&HDŽ$HR7'H[ 'H9XjH: 'H|HH=& 'HH$HGH5!'HHDHH$xH$H/5H$H$HDŽ$H9GwHt$HvufDH$H$HtH/H$HHDŽ$YH$H/H5'H$HDŽ$H9tHGH;$eHH/$HH$H/8H$H;|$HDŽ$H;=$6ÃHHPHHC8HDŽ$`H5'H=2'1j8HH$H!wH$H/uH7HDŽ$4'4'H4'bwfDH5'H 'H9HHy 'HXHH=e 'HH$PHGH5y$'HHIMH$H/)$IEHDŽ$LH5o''HH>HH$HU4'H5'H9pH'HHH='HH$HGH5%'HHIMH$H/#ID$H;D$HDŽ$dE1H;P$?H;$U#HH"H$HtHCHDŽ$H$IcHHLHDŽ$HT1Lt6HH$vH+#I,$"H$H;|$H;=$U!HHPHHt!HDŽ$H2'H5'H9pH'HHL%'MID$H5'LHHHHI,$HF2'H o'H9H`HN'H2HL5:'M|IFH5$'LHH}HH$>sI.FIEH5$'LHHHHH$HD$HDŽ$H9GMHOpH$IHtH/uHDŽ$HmMɽH$H/ID$HDŽ$LH5`#'HHHH$vI,$HD$H9CH$HoH$H$H/9H$HDŽ$H+THD$hHHH$vHIH;D$L;%k$pI$HPHI$2 H$HL$hHH$H$HHD$0HHH3H$H/2H$H$HH$H/2HT$hHDŽ$H$HHD$0HHH2HDŽ$H\$h\H;=$e.gH$HHPHHS褿IH;$-H H.'.' HDŽ$E11HX.'HDŽ$1H$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(-˾fDH$HD$hHD$X:fDH蘾H舾R{fDkfDHXHD$HH2f.Ht$pH|$P謾HH$κH;D$HH;$n ӃHHPHH/HDŽ$~H5'H|$P1?HH$~H;D$HH;,$ӃHHPHH.HDŽ$HD$XH;A$;H,'H5&H9pH&HHH=&HH$~HGH54 'HH~IMjH$H/.H5'H|$XHDŽ$2HH$H *$I9NHL$HLjH$H$H/-H$HDŽ$I.-Ht$PH$1譼HIʀH$H/-L;t$HDŽ$L;5$+IHPHIC-~H'H|$XHH$HGH5'HH~HH$K~Ht$H9pHDŽ$j6HPHH$U6H@HHH$H$H/,H$H6H$QiH$H$Ht H/uH$HDŽ$xH$H/T,Ht$XHDŽ$HHD$ HHH,H*'H&H9PxH&HxHL5&MyIFH5'LHHyHH$VyI.+HIxHD$HHIF HH$vHT)'HM&H9XvH,&HsvHL%&MuID$H5'LHHHHAI,$7+H5'H$H`HCH++H$H$L+HH$\H$H/+HDŽ$I.]+H$H/;+H$HDŽ$H5'HGHHHH$Ht$H9pHXHH@HHH$H$H/*H$HfH+HD$0uH4H|$0H$H/V*H$L$HDŽ$HD$@HD$81E1HD$(H$fHt$P1LQHHHQH;D$H;F$ DHHPHH$ E*H|$xH5'HGHHoOHH$NH|$PLۿHI@MHH$LH$LpHL$H9O#JH_HJHGHHH$H$H/H$H$H H+HD$ uH蛶H$H/IH|$ HDŽ$KH$H/MHDŽ$tImuL?H5X 'L軶HHLMH;D$H;$ DHHPHHEH^%'H 7&H9HLH&HLHH=&HH$&LHGH5'HHJHH$3JH$H/)H$HD$HDŽ$H9GIH$cIfDH$HtH/uLD$X LD$XMHDŽ$HH$H/HL$(HDŽ$Ht&HHD$XHHHuHLD$X訴LD$XI@H;$H;K$HXhHMH{MHH&LLD$(HLD$(HpLLD$(SLD$(HCSH@H;=$nܥH$HHPHHH;=ɷ$_&KH$HHPHHM蜳CH;=$y޸H$HHPHHgT]H;=9$薸qH$HHPHH H;$iHKZHL$HH5!'!'HDŽ$E1H!'HDŽ$E1HL$PH$1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$H @H <LHhHXLH[ӱ"H賱HD$Hf.HD$HWfH;q$ζ0{H$nL;%A$L蛶tHL$HHLl$XF ', 'E1H 'H\$hE1HL$PH$1HDŽ$HDŽ$1HD$HD$@E1HD$8HD$ HD$0HDŽ$HD$xHD$(HD$H&DkefD[fDKfD;DfD+TfDfDLL&fDLدq˯fDH踯H訯H9$HH;$H۴AH$HL$Ll$ z'E1HP'V'E1HL$XH$1HDŽ$HD$E1E1HD$xH;$HCAH$H)L$'' E1H'HDŽ$E1HL$XH$1HD$E1E1HD$xHXKfD;fD+`fDLt$0IFHXpH|H{|1訰HI{Hݱ$LHzLI*1*M\{LLSImHuL褭H8{H'H$H脫hH+HjDHXeLD$XFLD$X]@3fDH HVLHЬH賬裬lL萬H耬IPI@H\HLHLD$(HH$LD$(wGH+H$I(HD$(YH|$(HDŽ$H5A'HGHH GHH$FHH 'H&H9PkH&HWkHH&H kHCH5&HHL$HHjHL$HH$LcH+H$H5&H$HL$3HL$I*H$H/H$LLHL$HDŽ$@ HH$HL$bXI.I,$H$H/qHL$HDŽ$xHH$HL$fH$HL$HDŽ$HPpHH$HL$fH$H5+&HHL$.HL$#H$H$HHL$T HIHL$ mH) H$H/H$HDŽ$H/u訙HDŽ$HD$H$Ht!H;t$tH@@5B( HL$pHtHHD$HHHuHHL$vH)H5p&LLeHHD$pI.I,$H|$`LHGH@pHpH@HpHHzpH5'H|$HHL$蕐HL$H)uHzHD$HIfLXHH,L8QL(6HyHt$HHT$Ll$X0'E1H'H$1'Ht$PE1HT$`HDŽ$E1HDŽ$HD$1HD$@HD$8E1HD$ HD$0HDŽ$HD$xHD$(HD$HHD$hHD$pHt!H0HVHHuHHL$.HL$HtH+uHHL$HL$H$HtH/uHL$HL$H$HtH/uHL$͐HL$MtI,$uLHL$诐HL$MtI.uLHL$蒐HL$HtH)uHzH &&H=5&E1xBL;5A$jL蛕[HP&6&XHDŽ$1HDŽ$H&1H$E1HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(AH;=$ޔH$遹H;=Q$讔ʝH$H$LLT$8;LT$8@HLHD$HIfˎfD軎MfD諎fDL蘎b苎0fDLxHh[fDKGfD;fDL(L fD#fDfDۍ]fDLȍ軍fD諍OfD蛍fDL舍{sfDHh%[fDHHH85+L{Hj0&Y&Z#E111H&HD$H$fH$L$Ll$ HD$xHIH$HD$XH$HL$&&bE1Ht&HDŽ$1H$HT$XE1HD$E1E1HD$xH$HzL$L5&&~H&HDŽ$E1HL$XH$1HD$E1E1HD$xjL$H$HL|$@H\$8& 1H&&E1HL$XH$1HDŽ$HD$E1E1HD$xIWIGLDIM@LL轍HH$gI.2H$I/HD$( H|$xHDŽ$H5#&HGHH$fIMHD$I9GwI_HjMwHII/HT$HHLH$H+uHYH$aI.PH$H|$(HDŽ$H5`&HD$xHGHH͠HH$2ڍHIRHD$xHIFHHQH)&H5&H躐H$HLHH$WH$H/HDŽ$I.H+H&H ^&L$HDŽ$H9HUIH)&HaHH=&HH$aHGH5I&HH`HHW`H$H/)HDŽ$蓌HH$HIEH$Lh蜎HI.H $H5&HjH&H5D&H9p֛H#&HHH=&HH$HGH5&HHIM0H$H/SH5<&LLHDŽ$ŎI,$2H$LHHI H+IH$H/'HDŽ$I.ImHDŽ$HD$@1HD$8HD$ HD$0HDŽ$[LtLL؆LȆH踆gL訆M蛆*fD苆fD{fDLhLXImLE;fDH(H|B&(&HDŽ$HDŽ$1H&E1H$HD$E1HD$@HD$81HD$ HD$0HDŽ$E1GLHL$艅HL$rHHL$rHL$$L`[LL$OL$)HL$ LLLT$'LT$HLLT$@HT$oHT$LT$@HDZI. H* LT$@>HD$HD$ LT$@HH+l$0H9HHD$xHT$8L|$ Ll$@IHLp HD$0H)L$MM11E1LLzHL!HH2Ht[H9tfDHL!HH2HtH$Hp&j&HDŽ$1HE&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$E1HD$(Ht$XE1HD$x|H=&H͂&H5ւ&fH餶HpMŮ&&HDŽ$E1H&HDŽ$1H$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$ʭHro8& &HDŽ$E1H&HDŽ$1E1HD$E1XH5_&H=(&1HH$t]HH$H/u>HnHDŽ$& &HDŽ$E1H\&_Hnf& L&HDŽ$E1H&&)FI%H]n#& &@HDŽ$E11H&HDŽ$1H$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(H=&cHIHgȥ&_&g#E111H&HD$H$H=#&\I鑢H=&Hw&H5w& ]IoHfS& 9&HDŽ$E11H&HDŽ$1H$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(AH5&H=&1ۧHH$HH$H/uO5HL$HHeLl$XHDŽ$X&E1H.&4&T1HL$PH$1HDŽ$HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$h/HT$HHdLl$X&y&PE1H_&HDŽ$1H$HT$P1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hjHdآ&e&#E1E1HD$H&wHc&e&#E11E1Hi&HD$H$H;5$L3HD$H|$toLd$H\$I9L;%6$L;%6$ ك鼞HEc &Y&U#11E1Hӡ&HD$H$^Hc̡&Y&W#11E1H&H$(1LYHD$?L$H$Hbr&X&HDŽ$E1H2&1H$H\$XE11HD$E1E1HD$x颠HPH&5$H5L$H81M9mH!b&͠&HDŽ$HDŽ$1H&E1H$HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HaZ&e@&#1HD$H&H$鲟HZa &e&#1HD$H&H$xH a&e̟&#1HD$H&H$>8HL$H`&e&#HD$Hc&H$H=&HL$VHL$H閔H=&Hjq&H5sq&HL$VHL$Hj8HL$IɓHC` &e&#E11HD$Hʞ&H$^H=f&HL$,VHL$I7H=H&Hp&H5p&HL$?VHL$I H_&eh&#E11E1HI&HD$H$ԝH|$`)7H}H;1$L/HD$H|$uuHC_ &V&+#E111Hѝ&H$eH _ӝ&V&)#E111H&HD$H$&Lt$HD$I9L;5V2$ L;51$ B1LdUHD$?H^P&66&A!HDŽ$1E1H&HD$1H$HD$@HD$8HD$ HD$0HDŽ$HD$(HDŽ$HD$xIH=Q&H:o&H5Co&LT$HTLT$(H]&<t&{!HDŽ$1E1HL&HD$1H$HD$@HD$8HD$ HD$0HDŽ$HD$(HDŽ$HD$x釛H/]&eۛ&#E11HD$H&H$JH\&<&!HDŽ$HDŽ$1Hm&E1H$HD$HD$@1HD$8HD$ HD$0HDŽ$HD$xHD$(鱚HY\&<&}!HDŽ$HDŽ$1HԚ&E1H$HD$HD$@1HD$8HD$ HD$0HDŽ$HD$xHD$(H= &LT$QLT$H[k&6Q&VFH\$HHH]&C&HDŽ$E1H&HDŽ$1H\$PH$1E1HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(HD$H=HG&&Ll$XHu&HD$HHDŽ$E1HDŽ$1E1HD$P*H5j&H=&1聉HH$tNHB}&(}&HDŽ$E1H}&HDŽ$1H\$PH$1E1HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(HD$H"|H\$HH=|&q|&HDŽ$E1HK|&HDŽ$1H\$PH$1E1HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(HD$Hk{HGHH$JHWHHH$H$H/u~ H$HH$HbH$IHaJgJHt$HH<Ll$XY{&?{&E1H%{&HDŽ$1H$Ht$P1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$h0zH$kIH\$HH;z&rz&IHDŽ$E1HLz&HDŽ$1H\$PH$1E1HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(HD$HlyH\$HH;y&y&HDŽ$E1Hy&HDŽ$1H\$PH$1E1HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(HD$HxHT$HHX:y&y&HDŽ$E1Hx&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$E1HD$(HT$PE1HDŽ$HD$x xH=e&HM&H5M& 0H3H=d&/H3H\$HHv9Ll$X7x&x&E1Hx&HDŽ$1H\$PH$1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hwh3[HFHL$HH8Ll$XXw&>w&sE1H$w&HDŽ$E1HL$PH$1HDŽ$HD$E1HD$@HD$8E1HD$ HD$0HDŽ$HD$xHD$(HD$H&H5>&!H=6H4+L|$(i&/i& HDŽ$E1Hi&HDŽ$1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$HD$xHD$(hH*L|$(Ti&3:i&!HDŽ$E1Hi&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$HD$(HDŽ$HD$xJhH)h&h&FHh&=uH1HAYH)h&gh&HPh&HDH)Qh&7h&HDŽ$HDŽ$1Hh&HD$E1wH|$`Iq[H,)g&6g&8!HDŽ$HDŽ$1Hg&E1HD$鐬H(g&6g&3!HDŽ$1E1Hcg&HD$1H$HD$@E1HD$8HD$ HD$0HDŽ$HD$(HDŽ$HD$xfHT$HLMH=S&H9&H59&IvH(f&6f&.!HDŽ$11Hf&E1H$HD$HD$@1HD$8HD$ E1HD$0HDŽ$HD$(HDŽ$HD$xeH=R&IuH]'#f& f&HDŽ$HDŽ$1He&HD$E1H'e&e&HDŽ$1HD$He&HD$@1H$HD$8E1HD$ HD$0E1HDŽ$HD$(HDŽ$HD$xdH=Q&H+WH=Q&H8&H58&H WH>&e&d&HDŽ$11Hd&HD$1H$HD$@E1HD$8HD$ E1HD$0HDŽ$HD$(HDŽ$HD$xcH%fd&Ld&HDŽ$E11H$d&1H$HD$HD$@1HD$8HD$ E1HD$0HDŽ$E1HD$(HDŽ$HD$xWcH$c&c&HDŽ$1HDŽ$Hzc&1H$HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(bHa$'c&= c&"E1HDŽ$1Hb&E1H$HD$HD$@1HD$8HD$ HD$0HDŽ$HD$xHD$()bH#b&/}b& LHcb&HmGL|$(H$E1HDŽ$HDŽ$1E1HD$HD$@HD$81HD$ HD$0HDŽ$HD$xHD$(zaH"#a&/a& Ha&OH"L|$(a&/a& HDŽ$E1H|a&1H$HDŽ$1E1HD$HD$@HD$8HD$ HD$0HDŽ$HD$xHD$(`HE1E11L|$(H$HDŽ$HDŽ$1HD$HD$@HD$8HD$ HD$0HDŽ$HD$xHD$(+`HL$HH!Ll$X`&u`&E1H[`&HDŽ$1HL$PH$1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hf_H\$HH !Ll$X_&_&CE1H_&HDŽ$1H\$PH$1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$h^HI _&`^&#E11E1H^&H$j^H$H ^&^&HDŽ$1H^&1H$E1HD$HD$@HD$81HD$ HD$0E1HD$(Ht$XE1HD$x]HmHn4^&%^&1 HDŽ$E11H]&HDŽ$1H$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$("]HlH]&$h]&' HDŽ$E11H@]&HDŽ$1H$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(p\H5H@&H=!Z&1 `HH$t]HŞH$H/uHHDŽ$\& {\&#HDŽ$E1HU\&XH_\& E\&HDŽ$E1H\&"IeHVM\&[&HDŽ$1H[&HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$O[H=1H&HEdH=H&HF/&H5O/&H#dHM^[&D[&HDŽ$1H[&E1H$HD$HD$@1HD$8HD$ E1HD$0HDŽ$HDŽ$fZHMZ&Z&HDŽ$1HZ&HD$1H$HD$@E1HD$8HD$ HD$0HDŽ$HDŽ$YHt$HLMkH$`Hg-Z&Z&_HDŽ$E11HY&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$E1HDŽ$HD$x'YHMY&xY&HDŽ$1HSY&E1H$HD$HD$@1HD$8HD$ E1HD$0HDŽ$HDŽ$XHBY&X&mHX&HX&X&zHDŽ$E1HDŽ$HX&1E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1.XH|$(@&_HJX&=0X&!E1HX&HZ X&=X&!E11HDŽ$HW&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$HD$xHD$( WH$$3]|H$HoW&UW&HDŽ$E1H/W&1H$1E1HD$HD$@HD$81HD$ E1HD$(HT$XE1HD$xyVHHD$0/H$HV&V&HDŽ$E1HV&1H$H\$XE11HD$E1HD$@HD$81HD$ HD$0E1HD$xHD$(UH$ a.H=B& HN"Ht$HHKLl$X V&U&(E1HU&HDŽ$1H$Ht$P1HDŽ$HD$E1HD$@HD$81HD$ HD$0E1HDŽ$HD$xHD$(HD$HTII HT$HHLl$XFU&,U&#E1HU&HDŽ$1H$1HD$HD$@HD$81HD$ HD$0E1HDŽ$HD$(E1HT$PHD$HHDŽ$HD$x)TH|$`~H8 HT&kT&E1HQT&H$bLHDŽ$bH$aH$HDŽ$aHD$(H5>Q&HDŽ$HxXYf[HD$(HLHE11E11E1HpHt$HLl$X1H$HDŽ$E1HDŽ$HD$Ht$PHD$@HD$8HD$ HD$0HDŽ$HD$xHD$(HD$HHD$hHD$pRLwMe HGIHH$H$H/uH$H$L/:H$I.) L HR&R&E1HR&5H R&R&H=C5|R&wH|$(H$H$H$ՒH5F6&H=O&1xUHIt=H8I.uLHSR&Q& HQ&H,Q&Q&HQ&HQ&Q&HQ&HHL$0HHL$0H$I HQ&hQ&HQQ&H;# HH$Hm3Q&Q&HQ&餓HT$HHALl$XQ&P&E1HP&HDŽ$1H$H\$h1HD$HD$@E1HD$8HD$ E1HD$0HDŽ$HD$(HT$PHD$HHDŽ$HD$xOH$1jH=<&HH\$HHXLl$XP&O&!E1HO&HDŽ$1H\$PH$1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HNH=<&H#&H5$&IH=;&IHT$HHfLl$X'O& O&iE1HN&HDŽ$1H$1HD$HD$@HD$81HD$ HD$0E1HDŽ$HD$(E1HT$PHD$HHDŽ$HD$x NdHkH=E:&H#&H5$&HzH=#:&HfHo5N&N&E1HN&`Ht$HH6Ll$XM&M&N1HM&HDŽ$1H$HD$E1HD$@HD$8E1HD$ HD$0HDŽ$HD$(Ht$PHD$HHDŽ$HD$xLHMM&3M&YLl$XHDŽ$E1HM&HD$H1HDŽ$HD$P髗H4L&L&E1HL&pIHT$HH Ll$XL&L&*1HuL&HDŽ$1H$HD$1HD$@HD$8E1HD$ HD$0E1HDŽ$HD$(HT$PHD$HHDŽ$HD$xKIT$HH$I\$HHI,$uLHCIܽ1ҿA^H K&K&CHDŽ$E11HjK&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$E1HD$(HDŽ$HD$xJH=7&HN&H5W&H H=7&NH Ht$HH J&J&jHDŽ$E1HJ&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$E1HD$(Ht$PE1HD$HHDŽ$HD$xI u H HT$HH7 I&I&oHDŽ$E1HI&1H$E1HD$HD$@HD$81HD$ HD$0HDŽ$E1HD$(HT$PE1HD$HHDŽ$HD$xHH MI&3I&lHI&魓H` &I& I&rHH&醓H|$XH* H&H&5HDŽ$E1HH&鳚Ht$HH Ll$XH&H&E1H|H&HDŽ$1H$E1HD$HD$@HD$81HD$ HD$0E1HDŽ$HD$(E1HD$hHt$PHD$HHDŽ$HD$xGIHT$HH G&G&HDŽ$E1HG&1H$HD$HD$@1HD$8HD$ E1HD$0HDŽ$E1HD$(HT$PHDŽ$HD$xF-I HLIfDMLHPL|$(G&F&PHDŽ$E1HF&1H$HDŽ$1E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xFH=3&H&H5& HH=2&H{H#H8uH Ht$HHOLl$XF&E&E1HE&HDŽ$1H$Ht$P1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hDAHKE&1E&,HE&鼇HTE&E&=HD&鋇HGHH$HWHHH$H$H/uH$HHT$HH$+H\$HHLl$XD&kD&sE1HQD&HDŽ$1H\$PH$1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hHD$pSCH;S#7H|$pHHH$HT$HHLl$XC&mC&E1HSC&HDŽ$1H$HT$P1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$h^BH$H;|$H;=# H;=#v x!H$^ H|$p1{HHL$HHLl$X_B&EB&E1H+B&HDŽ$1HL$PH$1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$h6AH;6#HH$Ht$HHLl$XrA&XA&@E1H>A&HDŽ$1H$Ht$P1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hI@H=Q-&H&H5&MH H$F1H&H|$`ZgHt$HHLl$XZ@&@@&qE1H&@&HDŽ$1H$Ht$P1HDŽ$E1HD$E1HD$@HD$81HD$ HD$0HDŽ$E1HD$xHD$(HD$HHD$hHD$p(?Ht$HH?&w?&HDŽ$1HR?&HD$1H$HD$@E1HD$8HD$ E1HD$0HDŽ$HD$(Ht$PHDŽ$HD$x>HL$HH%Ll$X>&>&E1H>&HDŽ$E1HL$PH$1HDŽ$HD$E1HD$@HD$8HD$ HD$0HDŽ$HD$xHD$(HD$HHD$h=AWAVIAUIATUSHHhH-#L% #L #&HHD$0HD$XHFHl$8Ld$@Hl$HLL$P;HH|EHc H1LK8LL$PLC0LD$HHK(HL$@HS HT$8HsHt$0HL9H;=# ۉ$LHh[]A\A]A^A_1ILH1ILfD1IzfD1dfH~@H|$XDHCfDHH$L 'HZHH H5 LOH#HNH8LDH1H<&vx<&$$H\<&H H=-vTHh1[]A\A]A^A_H9HT$(Ht$ HL$LD$LL$YLL$LD$HL$Ht$ HT$(`HLL$LD$HL$Ht$ HT$(qH;&v;&H;&5;&59HHHD$HUCHcHHC@HD$XHC8HD$PHC0HD$HHC(HD$@HC HD$8HCHD$0LfHt$IHHCHcHH52&LIHVHHD$0MH5o#&LHVHtHD$8IMH5U$&LHViHtHD$@IM~rH5O&&LHVCHtHD$HIM~LH5I0&LHVHtHD$PIM~&H5#&LHVHt;HD$XIM-Ht$0HT$8HL$@LD$HLL$PH|$XHD$HL$HT$0LKH5%L_yH 9&v9&H9&fff.AWAVAUATUHSHHL%Z#HH|$HDŽ$L$[HFH] H Lv HkLt$HHD$`HD$pHDŽ$HHH9& L(hE111AHHAHHD$`HD$pHH|$`H/) HD$pH'&HD$`HD$pHD$HH8&H9PlH&HHH=&HH|$`|HGH5.&HHHH$H|$`H/0 Hr8&Hk&HD$`H9PHA&HHL5-&MHIFH5)'&LHHHH.I. HEH;#IA1E1H;y# H;<# T$ %HIŋT$ MtLpHt$HHcHH1HItLH&HKDJ:HHD$`ImF Hm H$HW#H9G^Ht$`uHD$pH|$`H/ H|$pHD$`H$H/b H|$pH;=|#HDŽ$H;=#{ HHPHH HD$pL9d$HD$HHh HD$ HHL8Mt M9 HPHuML`LhtIMtI$MtIEH6&H5&H9p)H&H;HH=&HH|$p6HGH5&&HHGHHD$`H|$pH/ HD$`H5#HD$pH9pHHt$OtH|$pH$HtH/uH$HHD$pH|$`H/W HHD$`HHD$`HHD$0HD$pHH$MHDŽ$HD$pHBHD$`HD$`HB tI/uL4MtI,$uL MtImuL H4&H5>&H9psH&HHH= &HH|$`HGH5@&HHWHHD$pQH|$`H/{H 4&H5&HD$`H9pMHy&HHH=e&HH|$`HGH5%&HHIMH|$`H/HT$pHD$`HBH;#HAE1H;#H;#HI,HD$`Ht IFHD$`HT$0IcLHHIT1H|$pOl5HH$I.H|$pH/H$HD$pHHD$8H$HDŽ$HPpHx HD$@H2&HT$(HD$HD$L%$&LLMuL>HHD$P H@HH4LLH|$PHHD$P HD$L%u$&LLMuLH HHLMLLHAHHD$p HHH; #LpMH@IHH|$pHD$pH/H|$pL$HGH; #H;#Z HWBG E1 LzuLgL--#P JA;MH A LLAIp AUNH BR9M LH$I.uL9H$HH|$pH/H$HD$pH/HDŽ$E1Ll$HD$XHLd$(HD$ I H|$:fDE1HfWBDLL$kL$CII9Xu LHNH&1H&%&,NE1HD$8HD$0H&HD$@1H; #H; \#HHQE1 LquL`IL-#p NA;MH _1LAI#P JH AE=P@9|MtPLH$@$11HVWH=Xt1H5&1H(HDuH#H5HD$ H:sHD$ H&%q&ME11HD$8HL&HD$0HD$`HD$@Hy?&%%&NE1HD$8HD$0H&HD$@YH1HAwff.AWAVAUATUHSHHL-#HH|$ HDŽ$HDŽ$L$i$HFHHHLf(H] HmHHD$pHDŽ$HDŽ$GHHD$)H&L(hE111AHHAHHD$07)HHH. He& L(hE111AHHAHHD$8"!HHH HD$8H5_&HXHH\$H&!HD$H|$HHp&f."!HD$0P M9ĦHD$(H fDHL8Mt M9HPHuMLpLhtIMtIMtIEH8&H%H9XH%Hf!HH=%HH$^!HGH5&HHo!HHD$pi!H$H/HD$pHҩ#HDŽ$H9X\!HLaUH$H$HtH/uH$HHDŽ$i!H|$pH/JH|$HD$p袩HHD$p芪HHD$H$HH$MHDŽ$HDŽ$HCHD$pHD$pHC tI/uL8MtI.uL%MtImuLH&H#%H9XH%HHH=%HH$hHGH5B%HHwHHD$pm H$H/HDŽ$jHH$ H\$HH$HXoHH$ H&H0%H9XH%HHL5%MIFH5&LHHIMI.H55&H$LLL$(LL$(HI)H$H$H|$pHI"H|$pH/H$HD$pH/H$HDŽ$H/pIGHDŽ$IAwI HD$PH&IHEtHEHPHHHcHMLHH$HH#H H5L 7HOH81LBH H&Mv&CCHZ&H H=dMRU1H[]A\A]A^A_MfW1۾H=%&"HD$ L-&HLLeLCHHD$@r"H@HH LHH|$@HHD$@U"HD$ L-z&HLLeLH!HHLMP LHHAHH$!HHH; Ϥ#HDŽ$`HPHH$KH@HHH$H$H/ L$H$ML$HGH;ޥ#H;#HWB#1 LbuHo葥x WP H#;"HLAHhx OH HФ#BR9 HX"H$Ht H/u HHDŽ$"H$H/ HDŽ$Hm 詧HD$XLHH|$HH~xHD$ Ld$PE1L|$PL|$HHLpHH LHD$(HD$IHHD$ HLIfDLD$H|$(HMLLHyH\$ L9uL|$PH|$XH\$@H5%1HzH;HWH|$HH H HHSHHuHIL|$(1E1E1LH|$8HHD$HHHu谟H|$(HtHHD$HHHu荟H|$0HtHHD$HHHujMtI,$uLVMtImuLBHtHmuH.H|$HtHHD$HHHu MtI/uLH]HcHH*]EHH Hڋ]H*]EHH H*虞Hߐ苞H5Z%H1 &M9 H &H%H9XH%HHH%H)HCH5H&HHHIMH+" ӡHHkI$L`HHD$p H4 &H=M%H9x H,%H HH%HH$; HPH5L%HHH HH$ H$H/Y H$H5%H|$pHDŽ$ H$H/ HT$pHLHDŽ$9HH$I. H+ H|$pH/J H$HD$p1H + &Ht$0HEL$HDŽ$LHL9H$HDŽ$IEH5.%LHH`HH$)H|$HHD$p ֟HH HD$pH$HHD$pHC8HHD$p9 H$H/2HDŽ$H+ HD$pH%HD$pHD$H &H9X H%H HH%HHD$p[ HPH5%HHH0 HH H|$pH/rHD$pОHHD$p H|$HHD$pHx۠HH$D H# &H=%H9x H%H HL5%M IFH5c%LHHHH$I.H$H5%H$'H$H/H$Ht$pHHDŽ$< HH$H+H|$pH/gH$HD$pH/lL$HDŽ$IH$H %HDŽ$HD$(HH@HD$XIEHD$@HD$ LIVHHT$`蝞HHD$PH@HT$`HHjLH|$PHHD$PHD$ H%LHIVHHT$`7H HHHT$`LMLHAHH$ HHH; #HD$pHPHHT$pH@HHH$H$H/Ht$pH$HnFH|$pH$HtH/uH$HHD$ps H$H/ZH$HDŽ$H/,HDŽ$膞H|$@HD$`%HD$ E1L|$hHXHH HH\$IHH\$ H\$XLd$XLd$HHl$HLMII0LD$MLLH0HH6QAFIF E1Ʌ-yH(H0H0H@(AE;N}NIcI4H0H@H0Pt8H(AHR8HcR H0E;N|HH\$ H;l$@:Ld$XHl$HML|$hH|$`FLt$PH5%1LHIHD$HHIH H+gIHft{y9fHD(H0H(H)0H0HcHH8ImuL܆H5%%)DH%H%%%DE1E1H%1HD$1E1E1LHͯH}H%j%ZFE1HP%FH=*%赬HHv<%"%EH %HD$pH@%%EH%XHPHH$H@HHH|$pHD$pH/u芅H$Ht>H|$pLZHĵ%p%EHY%HD$pH|$pHT%:%\FE1H %LfIt/4ItI,HF(H$HE H$HEH$HIII MMqH$H$L$,Hմ%%_FE1Hg%Hq%W%dFE1H=%H$LH5o%LH诱fHX%M%CCH%H5%HHV蠈HH$IH5%HHVxHH$t:IMH5'%HHVKH:H$IH=HU%M;%CCH%HEFHZ %%DHD$1E1H%E1vH"%%EDH%H=[%&HbH=G%H0%H59%CH@HŲ%q%vDE1E1HT%H^%D%sDHD$1E1H%E1IHS%%qDE1E1E1H%=H#%%mFE1E11H%E11E1jH%%DH%wH%g%DHP%H#HH8YHH%.%EE1E1H%LL|$(HM%%DE1H%wHs#LH8HDŽ$H%%FH%H\$@HHD$HHHuHiLE1E11E1E1*H#LH8qH`%F%FH/%Hv<%"%FH %HR%%CE1E1HD$H%11E1E1HD$0E1HD$8{H%%CE1E1E1H%HD$11E1E1HD$8/Hu41H`%F%GH/%Hs9%%FH%HL%%FH%BWH 1H=K:H5 %I}軤HL%%[DE1HD$Hh%1E1E1H1h%N%XE1H4%H (%B%H=5%% 0HY%%YDE1H%I@H$,HH5#H5ַH8~ufAWAVAUATUHSHHL-j#L5K%HH|$0HDŽ$HDŽ$L$L$:HFHHHHH$H[H=#H DH5'L HOH81LDHHH!%%GGH%H H=.1H[]A\A]A^A_DLv0HE(HD$(HEH] HD$HD$PHD$`HD$pHDŽ$HDŽ$HILt$PH-_%L9&%H׀#I9FH9EM9u!L9u LHLN|HI8GH;#@H;@#@(@I,$uL{FH|$PuS1HHPHHuqZ{j]A~ F} MIFH;E!H|$PH5%LW8H|$PHHPHHtHD$PfLyHD$H f.HL8Mt M9HPHuML`HhtIMtI$HtHEH%H p%H9HLHO%HLHH=;%HH|$`LHGH5b%HHYHHHD$pSHH|$`H/H|$pH|#HD$`H9GHH,(@H|$`HD$PHtH/uyHD$PHHD$`IH|$pH/0H|$PHD$pH/MHD$PtI/uLdyMtI,$uLPyHtHmuHHՃHPHH]HD$`LIH{HHD$`IHH8yHHD$pIH|$`H/CH|$pH;=|#HD$`@H;= |#@tsH/HD$p4HCw5HCHHbH[HcHfLl$(fDL`Hh]L9tF}I3H|$pH/wwmL9 }@3H|$`HHPHHwswcwSwHCwHD$ uHD$8HDHL8Mt M9 HPHuML`HhtIMtI$HtHEHG%H%H9XnBH%H:HH=k%HH|$`:HGH5%HH`:HHD$P!:H|$`H/cH|$PHx#HD$`H9G:Ht$z$f.H|$`HD$pHtH/uvHD$pHHD$`*:H|$PH/HL$HD$PH\$pHHD$HHHHCHD$pHH5.%HHC9HHD$p 9H5%H9yHPH;y#8HxuxW@H5ay#HHHT$PH0HVHHbH|$PH;=6y#HD$pH;=x#H7HVHH0HD$PqHCH5v%HHH7HHD$P7H5%HuHHD$p\7H|$PH/H|$pH;=x#HD$PH;=w#(HHQHH@HD$pAH%H5%H9p6Hr%H6HH=^%HH|$`[6HGH5U%HH66HH$5H|$`H/H5%1HHD$`tHHD$`5H\vHH$N5HHsHHD$P4H$H/pHt$PH|$`HDŽ$vHH$4H|$`H/%H|$PHD$`H/]H$Hiu#HD$PH9G3H$ fH|$PHD$pHt H/urH$HD$PH/H|$pHDŽ$^3H$H/H|$pH;=Uv#HDŽ$H;=u#^H7HVHHSHD$pD$MtI/uLqMtI,$uLqHtHmuHqD$2H%H %H9Hi2Hӻ%HE2HH=%HH$1HGH5%HH1HH$1H$H/~HDŽ$tHH$1HH$HXwHHD$p0HM%H5%H9p;H%H;HH=Ѻ%HH|$P~;HGH5%HHY;HHD$`:H|$PH/HT$`H5%H|$pHD$PJwbH|$`H/L$HD$`HT$pH$ID$LM:Ht$HT$sH-bs#H HT$Ht$;MH W:LAIsp UNH BR9{M4Ld$`H$H/KH$HDŽ$H/H|$pHDŽ$H/HD$`HD$pH+HD$H|$HD$`H5%HGHH>4HHD$`3HP3HPHmHsHJ 6HǐtII 6HD$`HHSHH&L9l$HD$``#HD$MLxUI11HDIHH)H9:HHL9uH:H-%I9Hr#I9FH9E  A~ :} %IFH;EfH-%I9 Hqr#I9FH9E5 - A~ c%} %IFH;Ef.H;\$ P$L9l$(H%H L%H9HR H+%H&HH=%HH|$`&HGH5.%HH&HH$"&H|$`H/LH$HMo#HD$`H9G%Ht$(H|$`HD$pHtH/uvlHD$pHHD$`&%H$H/H|$pH;=ap#HDŽ$H;=o# HHPHH HD$p HD$(H5n#H9p[$HH$L=lHHD$p$oHHD$`2HT$pH$HHD$pHPkHHD$p{2H$H/H|$`HDŽ$H/jHD$pHD$`HD$pHD$H%H %H9H1H%H1HH=%HH|$p1HGH5b%HHm1HHD$`41H|$pH/HD$pnHHD$p0HL$HHD$pHHpHH$0H%H5<%H9pP0H%H,0HH=%HH$/HGH5%HH$HHD$Pm$H$H/HT$PH5M%H$HDŽ$pH|$PH/ LD$`HD$PH$Ht$pI@HH4HL$HLD$@Ht$8HT$(umH-l#H HT$(Ht$8LD$@y;}HL$Hx {2LHD$(9mp UNHt$(H BR9?H1Ht$PH|$`H/H|$pHD$`H/ H$HD$pH/MHl$PHDŽ$HD$PoHEH51%HHHU1HHD$P1LhHH$1H|$PHXqHHD$p1H|$PH/H$HD$PH/hH|$pHDŽ$HG1HGHHnHR@0mHD$(H|$(90H|$pH/ L9HD$p/HEH5%LHD$81H%HD$H HL$0HLIVHHT$@lHI%/H@HT$@HHgLLHI /HL$0H%LHHD$HIVHHT$@-lH|.HHHT$@LMLHAHH$i.HHH; i#HD$PHPHHT$PsH@HHH$H$H/Ht$PH?H$dH|$PHD$pHtH/ueHD$pHHD$PH$H/H|$pHDŽ$H/HD$plHL$0Ht$8LLL$(LD$ HD$@H H4$HHLFdH|$@AIdH5ҷ%1LIMHQHIU$HHHQHHAoHEHHL$HHD$ HHHuHdHtHEHPHHUuHdHD$HtHHQHL$HHuHdH!CHHD$ H|$ `iHRHӔ%l%H1HD$H]%CHD$ CHd$ H D$ H\$ CHD$ CHD$ CHd$ H D$ H g#HHfHY%H %H9HOHq%H+HH=]%HH|$PHGH5%HHHHD$`H|$PH/o HCHD$PHH5%HHHHD$PjH%H5%H9p0H%H HL %MIAH5%LLL$HHLL$HH$`I) HT$`HBH;=e#IҹE11H;f#H;dg# t$@LL$L$DfHH$HcL$LL$Hct$@MtLHHT$PHH|$`HD$PHTH$HHDŽ$HT1LHHD$pH$H/6 HDŽ$H|$`H/ H|$pH;=e#HD$`H;=2e# HHQHH HD$p?D$jM9uL9uHLaHHH;3e#@H;d#@g@H*uH`HH9wcHHD$`H=-%HhHHD$pH|$`H/u`Ht$pH=%HD$`HHD$`6H|$pH/uO`H|$`HD$plH|$`H/u,`HHD$`B%(%FK1HD$H%S M9u3L9u&HLV`HHH;c#@H;Hc#@ @H*uH_Hɚ;H5%H=%1HHD$`HrH|$`H/u2_HHD$`H%.%J1HD$H %Y L`HhL92dH|$PH7HVHHD$^D$L9wc~H|$pHHPHHhk^HD$pcL^HHD$p**bHHD$H$y*HHD$(HHFHD$pHDŽ$HD$pHF ]fDH]]fDH]]MfD]xfD]cfDHt$0LHIVHHT$@bHIH@HT$@HHLLHIHL$0H %LHHD$HIVHHT$@kbHfHHHT$@LM^LHAHH$SHHH; T_#HD$PHPHHT$PH@HHH$H$H/7Ht$PHkH$ H|$PHD$pHtH/u8\HD$pHHD$PnH$H/H|$pHDŽ$H/HD$pbH|$0IHD$8LL$(LD$ HLLH H$cLZH5%1LI]HSHIUUH\'HHSHHRHg[EfHX[>HH[;[fD+[fDD_@$wfZ fDZ3fDZ fDZfDHt$(^Ht$(@$ZfDZfD{Z)fDkZfDE1D`D`PII LDL9r_5&H|$pHHQHHD$YD$YfDYafDHh^HD$ fDYfDYY`sYHD$(DGHD$(GHd$(H D$(DGHD$(INHUH9tHt HAv DE D@9@ = I~0MNH@IDA  Hu0HHA@HDN u DA9oH H{[H|$P@@INHUH9tHt HAv DE D@9@ I~0MNH@IDA eHu0HHA@HDDA9yH1HZ@fDINHUH9tHt HAv DE D@9@ I~0MNH@IDA Hu0HHA@HDDA9YH1HY@dfDL1hL9F(\Z H|$p2L9\"H|$pIEHH$HH$IE9[V'fDKVRfD;V'fD+VfDV`fD V5LUIUUfD8UfDUfDUfDD$UD$U@JUfDvUfDD$bUD$IB@HD$PL$HE1H$H$H$IRHBR uMZLL$HLLL$HHD$pLH|$PH/JH$HD$PH/%HDŽ$TLTHHD$`D\XHHD$HD$pHHD$`HD$pHD$`HFLHD$ )THD$ LHD$ THD$ HT1H$9H$9H)%%rJH\$1HD$H%H|$PHt H/uSH|$`Ht H/utSH|$pHt H/u_SH$Ht H/uGSH$Ht H/u/SH 8%R%H=s55%1.H|$DHQ%%L1H%1M92LX'L9HHT$XHT$L9HHT$WHT${HD$PHHLL$LL$H$H$H$HHD$pLL$JH|B%t(%]IH%H|$PLL$LL$HD$PLH$H$HDŽ$H|$`HDŽ$H|$pHD$`mLt$8H5%HD$pI~XkH p%%H=5m%hH$H$Ht$pLH|$pH$HD$pH$HDŽ$HD$8HLLHDŽ$HH\$HSHH$H=8%HXHH$KH$H/ulPH$H=ͼ%HDŽ$HH$H$H/u(PH$HDŽ$?H$H/uOHUHDŽ$%|%OJ1HD$HӾ% fDHuHI~HcHuHI~HHuHI~H}OkODHDDpDdDDHw=%y#%IH %HD$8HLL1HH\$HD$/H|$P1SV5Hֽ%t%:IE1H%H=F%uHH=2%H%H5$%.uHH~v%u\%KIHE%/;VLL$YHz~@%u&%IIH%H=%~tIH=%Hh%H5q%tIH~%tɼ%?IE1H%UTH}%t%{%%J1HD$Hȹ%H {ҹ%%MH%rH5O#Ht$HH8JHDŽ$Hz%q%LHZ%+HN#Ht$HH8aJHzP%6%LH%lHcz)%%K1HD$H%:H1z%rݸ%IE1Hø%LJMHBIHH|$`HD$`H/tILT$`1ҿIBHy%tk%IE1HQ%;LL$2ILT$`1ҿLL$IBIҹ1uH5%H=W%1@HHD$`tSHH|$`H/uHHyHD$`Է%%fK1HD$H%Hx%%bK1HD$Hf%HiFHx`%F%"K1HD$H$%qL'FHF#HCx %%J1HD$HͶ%Hx׶%%K1HD$H%H|$(NHH$Hw%p%K1HD$HN%HwX%>%K1HD$H%iHGHHD$`[HWHHH$H$H/uFHt$`HHT$(H$Hwȵ%%K1HD$H%Hv%|%K1HD$HZ%H=%lH*%HuH%.%AK1HD$H %YHPu%%6K1HD$Hڳ%'Hu%ʳ%4K1HD$H%Ht%j%H1HD$Hv%Ht%hf%H1HD$HD%HGHHD$p#H={%HSLHHD$`H|$pH/uCHt$`H=I%HD$pHHD$ptnH|$`H/uCH|$pHD$`H|$pH/uCHsHD$p%k%H1HD$Hh%Hsr%kX%H1HD$H6%Hzs@%k&%H1HD$H%QHHs%k%H1HD$Hұ%HlHD$ Hsʱ%`%G1HD$H%LvIHHJcHHrz%`%K1HG%HrQ%7%K1HD$H%bHF0H$HE(H$HE H$HEH$H@II~uIIu-M~1H5%HHVqFHH$IMH$H$L$HD$(H$HD$MuH5%HIHVFHH$tZH5Ĝ%HHVEHH$tCIM{H5%HHVEH.H$IHEH=oy]Hp%%GH%5%H$LyH5z+%LHmHpf%L%GH5%LHtp:% %kJH\$1HD$H%FH=p%%fJH\$1HD$H®%Hp̮%%cJH\$1HD$H%G+Ho%q%aJH\$1HD$HJ%H=%eHH=ښ%Hc%H5l%eHH\$6HNo%w%IE1H%HGHHD$PHWHHH$H$H/u>Ht$PHtnH$H$Hn%ws%IE1HY%CHnc%wI%IE1H/%H$WHfn,%w%IE1H%HLHڤ%'Hf%ʤ%SLH%He%%oJH\$1HD$H|%H=poHt$HT$8HT$Ht$LoHcH1-HHtXH]Y%%QH֗%5%oGLt$0HH H1 f&HHHӏ%H=%HHCH%HL5˂%0HI;].HI*E1LHHL-ImIuL (MH+ID$L-L%LHLHHHEHHI,$H*HIH=%H/HHI,$HH *#H9ClHHIIHmfM I,$BLZImHW>%$%E1H %H!%5 %H=|HtH+uH&MI$LE1@oGHH HC+HL&Lp&LHD$[&HD$&LHD$C&HD$fH(&+L&bH&GL%L%H%L%LLfIt(/ItI(HF(HD$0HC HD$(HCHD$ H#III"MML|$ H|$(IH;i)#L|$H;'*#HKAE1 HiuLc)L-(#P JA;MH LLH(AUH H BR9|HIxf(@$Ht$HHBHHwHH HHHH OHG$BHTa%G%ϐ1H.%I,$uL$E1H % H;!(#Lt$ H;(#IT$B?HRE1 HT$uM|$'L-8'#P A;UP LLHD$I'P JAUH BR9|MLM]'@$Ht$ L>+HHSE% +%IE1E1H%11՟L͟LIŸH %%H=~5%H8S% ȑ%H%HHS%%E1H%HuH55|%HHV9'HHD$(IH5{%HHV'HHD$0IHR<%"%H %LsMLcII$H+HLL4xI.IfL!YHRɐ%%H%eH,&#H8T#I,$HQ%l%HU%uL;!H D%^%H=5A%E19k)H6H|QI#% %1H%HPQ%%E1HƏ%HH#Q͏%%H%HML;LH?HHP%e%sME1HH%5HPR%8%]H!%(IHtP%%[E1H%H'I0#'fD\H"#H5xE1E1H8 HP% %PH}%jHt$H辐H#HuH!#H5^YH8& 1LL8>H=Y"1C@{#HuHg!#H5YH8L4FOHH=X?"lkHt$ LHHH!#HUH5.UI1H81%HNj% P%H9%H*HZ!#H$L QH PHVH5.]AH81f%H[N%%CHԌ%HT$ LVH5 %LHIKBHN%%GH%H #H$L QH PH(VH5\AH81$HMV%<%=H%%JH}H}DH*Hf(L$L$ʀf(YX%^|YYf(^\^ \YXHXf.HHH?P H*YsHfffff.HGH?AUIATIUHS1HH~fDH}UADHL9uH[]A\A]fffff.AUIATIUHS1HH~-fDH}U H*YADHL9uH[]A\A]fffff.ATL%UH-SHH H;SHH H*HHcH;DAY sH f([]A\ÐHAL$HcH;\$\T$SL$$f(fW%f(ET$\$Y$Xf.OL$H []f(A\H;S :z\f(f(H []\A\fff.AUIATIUHS1HH~fDLAHH9uH[]A\A]fffff.AUATL%%UH-SHHH;S *Hc;DAY sH([]A\A]@H}L$ HcH;$\d$\$SL$ A(A W~II*\$d$YpYX.JL$ H[]A\(A]H;S ;H*Y *\(H[]A\\A]AUIATIUHS1HH~fDLAHH9uH[]A\A]fffff.AUATIUSHH~BHI1fDH}U x\f(dfWLADHL9uH[]A\A]ÐAUATIUSHH~_HI1fDH}U  wH*YZ\f(ffZWADHL9uH[]A\A]@AVAUL-ųATL%UHSHH @HL$HcH;\$\T$SL$%$Yf(YT$\$Y$L$Xf.wYXZf.w8H;SAA DH*HcAYtWD;l2(H[]A\A]@H;S -H*Y\(H;YT$S 5H*Y\(JWcT$(YX.yaAXQW)H[]A\A]fffff.AUIATIUHS1HH~fDLAHH9uH[]A\A]fffff.SHH0f.HsD$1fW|$f. = sf.|$|$ֿ\=ƿY|$(Qf.%r^d$@HL$fWYX rf.sf(D$H;YYL$ST$ Xf(5trYYY\f.wZT$ D$D$T$  |vYY$r\T$XYT$(Xf.L$4D$(YD$H0[f.D$qL$ \^D$-t$L$ f(D$Yf( q^\T$\$\f.sH;SHD$@ `qT$\L$f.b BqD$f(^L$\$f.rH0[fWH0[H0[kf(fSHH .UD$hW|$.I=/.|$N|$\=&"Y|$Q.&%^5߼t$d$HL$YWX .s(D$H;YYL$ ST$ (H*Y 5pYYYD$\.wYT$D$D$ T$ KYY\T$ XYT$X.L$,D$YD$ H [Wt=.|$5ʻt$@H;S HH*YT$T$  \L$T$ .r5 ^L$D$ (Y\$ .rifD$F\L$^D$t$(L$D$ Y( ^\T$ \$\.#H [H [x(HHH?PHHff.HHH?PHfff.HHH?PHHff.HGH?f.f(D„f.mD„{USH(-}f.Cf(1f(l$5smY\$ Q$^f(YX @Y\ <YX 8Y\ 4YX 0Y\ ,YX (Y\ $YL$ML$$X *q\$f(l$\%qf.^YX ޹X\rHH~Cf.\lL$Hf($L$H9$\}H(f([]Ðf(\H,H*XDfWff.HL$$L$YX$HHD$ YD$HfDHHH?L$$PL$YX$H@HL$ L$HYfHL$ L$ HYfSHH 5HkD$f.L$s>D$H` $HD$M  $H [X^f(fDf.rf.H;SH;$S j$^L$D$f(  jT$^L$$f( X$%jf.rfWf.v<$H [^f(ff.HYn HXfDSf(HH\$ $ $Hf(f(Y$f\$$HY[^f(fSHHHD$L$H[^f(ÐHD$q ^D$V \iHff(HfWf.zu HT$%  EiT$H^ f.HD$  ifW if(^L$H\f(= ffff.SHH$L$ fWf.wFH;Sf.lr ~\\f( YD$$H[\f(@X YD$X$H[fffff.SHH$L$DH;S"hh\f.f(vo  WfW^ YD$$$H[\f(DSHH$L$DH;SfWf.v g\^ YD$X$H[fH H>fffff.HHH?D$P Mg\f( YxQf.zD$HYf(fff.SHHD$ L$H$Y 'kf(L$L$f(Qf.zQf.z.$H[Y^D$f(\$f(f(T$T$f(뷐USHHXf.D$sf.zuHX[]1t$fWf( fD$H $H;S $Yf.L$wHXH[]fDQf.D$ $ $D$8Y 5=X ߲f(YL$\ Q\ٲf(5ղ\-^l$@˲^X5t$H\|$0f(X|$ H;Sf(H;\Ki$S$D$f(,ifT$e\D$ ^$XD$YXD$XID $H,f.:d$r|$0f.Ha-f.v f.If(\$(D$D$H\$(t$@Y$D$^XH*HEL$X $\H*YT$8L$\T$$$L$\f.5 f(Ȑff.cSH\^f(CH[f.AWAVAUIATIUSHH$t H9r; =cMe$AEf(A}\f.t$X$t$P|$Xf(I*AM|$PT$A} YX\$AM(f(L$I \$H,Y\$PL$T$Qf.\$pIm0? \$PYT$YL$\ f(H*=fL$XT$|$(fD(t$Au8f(XXɯf($f(Au@5AXA\^f(l$x\d$0AeHAmPX5fD(t$@AuXt$XYf(f(\f(^f(Yf(f(XYfA(AXf(D$`AE`f(\L$PYfA(^^f(fA(YXXYAY^D$hXAEhT$AUpt$8AuxXl$ AMI)IEH$DH;SL$ H;YL$SL$f(f.L$^f.L$bt$f(|$@H*\f(]`Y^XXD$0\X\$(fT`^\f.gT$Hd$L,d$T$HMI)LH?HL1H)H~D$pH*YD$(\f.\$XID$I9^\$PH*YHUI9|(fDH*Hf(I9^f(\Y}f.@$M)f.|$(MGH[]A\A]A^LA_f(L$X\|$P3@f.L$8L$HT$wto^D$`XD$0L,T$L$HLH?fWf.D„f(\d$Y^Yd$`^D$h|$x\f(L,T$L$HM9fWf.D„f(\d$8Y1^Yd$h!fDIGH9DH*Hf(H9^f(\^~f.of(H|$pH^f(H^XYH*X{^XD$(Yf(XT$H^f(L$L$f(T$Hf(\f.Xf.tIGH*$fD($H*HEDY|$HL*ID$L)D$H*$fA(D$fE(^f(f(l$YEYY$D$$|$H$f($^D$|$X$Y|$D$Hd$PYf(^I*$D- Y$D$D%DfA(DD$D$$Xd$($$$Yd$HXI*YfA(A^XfA(\f(fA(A^\f(A^\if(A^\f(fA(A^D MA^XfA(A^\f(fA(A^\f(fA(A^\f(f(A^\f($^fA(A^XfA(D^^E\D^\f(fA(^E\D^\f(fA(^E\D^\f(f(^A\^D$\f(^A^A^XXf.y@YT$$\XL,f.BzHj0f(|$Xz f(|$Pz8|$z@$zH|$0zP|$xzX|$@z`|$`zh|$hzp|$zx|$8|$ H*YY|$p=]|$(ff(T$T$L$DAUf(IATIUHSHHt H9rH*Im%XAEA]\\$ Ae f(d$L$L$Yf\$ L$f(d$D$YAEf(AUXYX5bXQf.f(M=YXf.H,I]0\$I<$d$AT$t$1d$f.f(\$w/eH\H)HH*YYH*Y^f.v8HPH9}\$I<$d$AT$L$1d$f.\$wHH[]A\A]f.f(z@YXH,;@f.BlfzHZ0b |$f(|$(\$ T$d$|$(\$ T$d$f(l$8\$0T$(L$ t$Tl$8\$0T$(L$ d$t$cff.HSHt|fWf.Duj Zf.r$H*Yf.rM[ PVh\H*Yf.f(s)tH)H[@1[@[Zf.+H)H[SHH$f(L$TupL$f. {pU$f.vu\HL$f(H$zL$f(Qf.zlXYX$H[H[fu$HH[@Y YHf(4HHH*X$H[zf(\${\$zSf(HH $f(\$$$Hf(f(Y$"\$$HY[^f(Df(Sf(Hf(H XL$T$^d$MT$L$f(Y }YYf(YYXQf.z\\L$T$H;YXL$ST$L$f(f(X^f.s Y^f(H f([f(T$\$2T$\$@SHH@D$0f(L$Sf.D$t$f.=OS|$ ^X|$(@f(T$L$^f(XSL$fWT$\f.syH;SY#L$(H;f(YXXT$ ^\T$YL$L$SL$fW\T$Y\f.HT$H;ST$D$f( Vf.L$v  fWT$0}RXsf(fTX SL$D$u\$8D$8f.fW\f.H@[fDH@[f.H;SX\QY֟H@[=QY|$ YXQf.znL$ Xf(XQf.|$\f(X^f(YXX|$ ^|$(D `H@[fWf(mfTT$ !XxT$L$2L$f(_SHH0%PD$ \f(D$(H;Sf.D$ H;D$SYD$(NPL$\f(Yf.r~f(L$T$T$D$f(t\$^OXH,L$H[fWf.DфEH0[fDf.r fHHH?OD$\$PL$$f.f(vfYHXf.wHfHHH?$PO\f(y OD$\$f(]L$^f(zHH,f.sff.Sf(HH \=N|f(|$D$H;SH;$S wN%_N^L$D$\$$f(f.XNf(w5*Nf.wf($L$^X`f(L$\M$l$Y^Yf(\M^f.EH H,[ff(H8f(H\T$(H?\L$ D$f(l$d$^4$P4$d$f.l$\$L$ T$(s7?M\f(\YYQf.zA\H8f(@YYQf.z XH8f($_$f($J$AT1HUHSt`HHHH HHH HHH HHH HHH II I ĸH9w"f.H;SD!H9r[]A\DH;SL!H9sH;SL!H9rf.AWAVIAUATIUHSHHt.HIH9w4EH;SD!A9rL4(HL[]A\A]A^A_@HEt!H;SL!I9sH;SL!I9rfLjH?SII9IIv)L1HII9IsDH;SII9IwMIn@JH?L$ SL$ AAMD9v*D1A9AsH;SAME9wI If.H?SIfAWAVAAUAATUSHt'Ht~EAt.fDH;SD!9rF4(HD[]A\A]A^A_DDzH?SEIA9v!1A9AsH;SIA9wH Et @H?SADAWAVAAUATUSHfH\$PtcfILtgEAAA+EPDUf#fA9s%uI?AWEDf#fA9rDH[]A\A]A^A_ÐA+A)DfH[]A\A]A^A_fADj+AHA ;EAfA9vR™Af9AsA"DEHMAfA9vuI?AWEH[]A\A]A6A^A_f.H?AWE-T$ H?AWT$ EAfDAWAVAAUATUSHH\$Pt`ILtiEAAAD+EPDU"A8s#uI?AWED"A8rDH[]A\A]A^A_DA+A)DH[]A\A]A^A_f.ADb+AHA DDAA8ĉvUE™A@8As@DEHMDA8ʼnvuI?AWEאHf[]A\A]A6A^A_fH?AWE-T$ H?AWT$ EAfDSHH\$ uH[f.At+A)H[HLL$H?QLL$AfAWAVAUMATUHSHH(HHt$HLHAVt$I9v$D$19щsDI>AVI9wH HL$KLII9uH([]A\A]A^A_ÐHE2E1HHLbH\$)DI>AVIL9IIs-HD$1II9HsfI>AVIH9IwLHD$KDII9uHH HHH HHH HHH III HE1fI>AVD!9wHD$KDII9uH([]A\A]A^A_HK1DI>AVHD$IDHH9u%H11HHH HHH HHH HHH HHH II I HE1@I>AVL!H9rHD$KDII9uC9/%Hffffff.AWAVAUMATIUSH҉t$HLHʃHH؃H9HGH D$HAE-HAEHAEHAE HAEAEI9toMI)LHH<Ht,fnL$IL1fpHHfAH9rHI9t)t$HPI9At~HAtI9~ T$ATH[]A\A]A^A_IEuuBE1Hɉ\$ D$~I>AVt$H9v$D$ 19щsDI>AVH9wH L$CLIM9uH[]A\A]A^A_ÉHHH HHH HHH HHH HH H-E1@I>AV!9rD$CDIM9uH[]A\A]A^A_H1I>AVD$ADHL9uH[]A\A]A^A_Hu;11A2( HfH_LȃHH؃H9HGH HH+HfA1}HfAqFHfAqAHfAq<HfAq7HfAq HfAq HfAqH fAqfAq H9II)MINMt1fnI>> from numpy.random import Generator, PCG64 >>> rg = Generator(PCG64()) >>> rg.standard_normal() -0.203 # random See Also -------- default_rng : Recommended constructor for `Generator`. Gets the bit generator instance used by the generator Returns ------- bit_generator : BitGenerator The bit generator instance used by the generator (   d T,н +p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-p-+p-p-+p-p-p-p-p-**p-p-*p-p-***p-p-*p-p-p-p-p-p-p-p-p-p-p-p-p-*P*+p-++p+`+p-p-P+p-p-p-@+0+p-@+@*++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++P*++`*+++++**++*++***++*+++++++++++++**++@+`+++++ ++++0+p*+0+{---{------{-{---{---H+H+{-----------------]-R-h--h-h-]-]---]----]-]--]-,<,<,,<,<,<,<,<,+,<,<,*<,<,***<,<,<,<,<,<,<,<,<,<,<,<,<,<,<,<,,,+<,+$,+,<,<,*<,<,<,,*<,,---------h-H---,--,,,------------------,-H-x-h-H---,----,--l***l******)\***\***)))****************l*l*,**D*T+)\***\****l*)*l*V,j,j,V,j,j,j,j,j,L,+j,j,+j,j,+++j,j,j,j,j,j,j,j,j,j,j,j,j,j,j,j,V,V,+j,+`,L,+j,j,+j,j,j,V,+j,V,/\,\,\,\,\,\,\,\,\,J,\,\,J,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,J,\/\,\,\,\,\,\,T.\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,,.\,D,.\/ -.\, -\,\,\,\,\, - -\,\, -\,\, -\, -\,\,|-\,\,\,\,\,L-\,\,\,.\,\,\, - - -\, - - - -\,\, -\,\,\, - -\,,\,\,\,\,,0\,\,\,\,/ x ,\@@@@@dH|HTHDlNQQQQQLܧܣ[RI@7. BLccacbdcXbXd`d$ll\l 0dĐԓ6)ܺ8Xh9999999:<:a::uuuuuu@Bk\$4DؼϼƼGq$ $$$FFFFFa_[[_Unsupported dtype "%s" for standard_gammaUnsupported dtype "%s" for standard_normalFormat string allocated too short.unable to allocate shape and strides.sum(colors) must not exceed the maximum value of a 64 bit signed integer (%d)probabilities are not non-negativenumpy.core.umath failed to importno default __reduce__ due to non-trivial __cinit__negative dimensions are not allowedmethod must be one of {'eigh', 'svd', 'cholesky'}method must be "count" or "marginals".mean and cov must have same lengthgot differing extents in dimension %d (got %d and %d)covariance is not positive-semidefinite.cov must be 2 dimensional and squarecolors must be a one-dimensional sequence of nonnegative integers not exceeding %d.check_valid must equal 'warn', 'raise', or 'ignore'both ngood and nbad must be less than %da must be 1-dimensional or an integerWhen method is "marginals", sum(colors) must be less than 1000000000.Unsupported dtype "%s" for standard_exponentialUnsupported dtype "%s" for integersUnable to convert item to objectProviding a dtype with a non-native byteorder is not supported. If you require platform-independent byteorder, call byteswap when required.Out of bounds on buffer access (axis %d)Non-native byte order not supportedInvalid mode, expected 'c' or 'fortran', got %sInvalid bit generator'. The bit generator must be instantiated.Insufficent memory for multivariate_hypergeometric with method='count' and sum(colors)=%dIndirect dimensions not supportedIncompatible checksums (%s vs 0xb068931 = (name))Generator.standard_normal (line 978)Generator.standard_gamma (line 1152)Generator.standard_exponential (line 418)Generator.standard_cauchy (line 1636)Generator.noncentral_f (line 1410)Generator.noncentral_chisquare (line 1556)Generator.negative_binomial (line 2953)Generator.multivariate_normal (line 3456)Generator.multivariate_hypergeometric (line 3811)Generator.hypergeometric (line 3232)Format string allocated too short, see comment in numpy.pxdFewer non-zero entries in p than sizeEmpty shape tuple for cython.arrayCannot take a larger sample than population when 'replace=False'Cannot create writable memory view from read-only memoryviewCannot assign to read-only memoryviewCan only create a buffer that is contiguous in memory.Buffer view does not expose strides zipf(a, size=None) Draw samples from a Zipf distribution. Samples are drawn from a Zipf distribution with specified parameter `a` > 1. The Zipf distribution (also known as the zeta distribution) is a continuous probability distribution that satisfies Zipf's law: the frequency of an item is inversely proportional to its rank in a frequency table. Parameters ---------- a : float or array_like of floats Distribution parameter. Must be greater than 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Zipf distribution. See Also -------- scipy.stats.zipf : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability density for the Zipf distribution is .. math:: p(x) = \frac{x^{-a}}{\zeta(a)}, where :math:`\zeta` is the Riemann Zeta function. It is named for the American linguist George Kingsley Zipf, who noted that the frequency of any word in a sample of a language is inversely proportional to its rank in the frequency table. References ---------- .. [1] Zipf, G. K., "Selected Studies of the Principle of Relative Frequency in Language," Cambridge, MA: Harvard Univ. Press, 1932. Examples -------- Draw samples from the distribution: >>> a = 2. # parameter >>> s = np.random.default_rng().zipf(a, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy import special # doctest: +SKIP Truncate s values at 50 so plot is interesting: >>> count, bins, ignored = plt.hist(s[s<50], ... 50, density=True) >>> x = np.arange(1., 50.) >>> y = x**(-a) / special.zetac(a) # doctest: +SKIP >>> plt.plot(x, y/max(y), linewidth=2, color='r') # doctest: +SKIP >>> plt.show() weibull(a, size=None) Draw samples from a Weibull distribution. Draw samples from a 1-parameter Weibull distribution with the given shape parameter `a`. .. math:: X = (-ln(U))^{1/a} Here, U is drawn from the uniform distribution over (0,1]. The more common 2-parameter Weibull, including a scale parameter :math:`\lambda` is just :math:`X = \lambda(-ln(U))^{1/a}`. Parameters ---------- a : float or array_like of floats Shape parameter of the distribution. Must be nonnegative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Weibull distribution. See Also -------- scipy.stats.weibull_max scipy.stats.weibull_min scipy.stats.genextreme gumbel Notes ----- The Weibull (or Type III asymptotic extreme value distribution for smallest values, SEV Type III, or Rosin-Rammler distribution) is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. This class includes the Gumbel and Frechet distributions. The probability density for the Weibull distribution is .. math:: p(x) = \frac{a} {\lambda}(\frac{x}{\lambda})^{a-1}e^{-(x/\lambda)^a}, where :math:`a` is the shape and :math:`\lambda` the scale. The function has its peak (the mode) at :math:`\lambda(\frac{a-1}{a})^{1/a}`. When ``a = 1``, the Weibull distribution reduces to the exponential distribution. References ---------- .. [1] Waloddi Weibull, Royal Technical University, Stockholm, 1939 "A Statistical Theory Of The Strength Of Materials", Ingeniorsvetenskapsakademiens Handlingar Nr 151, 1939, Generalstabens Litografiska Anstalts Forlag, Stockholm. .. [2] Waloddi Weibull, "A Statistical Distribution Function of Wide Applicability", Journal Of Applied Mechanics ASME Paper 1951. .. [3] Wikipedia, "Weibull distribution", https://en.wikipedia.org/wiki/Weibull_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> s = rng.weibull(a, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> x = np.arange(1,100.)/50. >>> def weib(x,n,a): ... return (a / n) * (x / n)**(a - 1) * np.exp(-(x / n)**a) >>> count, bins, ignored = plt.hist(rng.weibull(5.,1000)) >>> x = np.arange(1,100.)/50. >>> scale = count.max()/weib(x, 1., 5.).max() >>> plt.plot(x, weib(x, 1., 5.)*scale) >>> plt.show() vonmises(mu, kappa, size=None) Draw samples from a von Mises distribution. Samples are drawn from a von Mises distribution with specified mode (mu) and dispersion (kappa), on the interval [-pi, pi]. The von Mises distribution (also known as the circular normal distribution) is a continuous probability distribution on the unit circle. It may be thought of as the circular analogue of the normal distribution. Parameters ---------- mu : float or array_like of floats Mode ("center") of the distribution. kappa : float or array_like of floats Dispersion of the distribution, has to be >=0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mu`` and ``kappa`` are both scalars. Otherwise, ``np.broadcast(mu, kappa).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized von Mises distribution. See Also -------- scipy.stats.vonmises : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability density for the von Mises distribution is .. math:: p(x) = \frac{e^{\kappa cos(x-\mu)}}{2\pi I_0(\kappa)}, where :math:`\mu` is the mode and :math:`\kappa` the dispersion, and :math:`I_0(\kappa)` is the modified Bessel function of order 0. The von Mises is named for Richard Edler von Mises, who was born in Austria-Hungary, in what is now the Ukraine. He fled to the United States in 1939 and became a professor at Harvard. He worked in probability theory, aerodynamics, fluid mechanics, and philosophy of science. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] von Mises, R., "Mathematical Theory of Probability and Statistics", New York: Academic Press, 1964. Examples -------- Draw samples from the distribution: >>> mu, kappa = 0.0, 4.0 # mean and dispersion >>> s = np.random.default_rng().vonmises(mu, kappa, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.special import i0 # doctest: +SKIP >>> plt.hist(s, 50, density=True) >>> x = np.linspace(-np.pi, np.pi, num=51) >>> y = np.exp(kappa*np.cos(x-mu))/(2*np.pi*i0(kappa)) # doctest: +SKIP >>> plt.plot(x, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() unknown dtype code in numpy.pxd (%d) uniform(low=0.0, high=1.0, size=None) Draw samples from a uniform distribution. Samples are uniformly distributed over the half-open interval ``[low, high)`` (includes low, but excludes high). In other words, any value within the given interval is equally likely to be drawn by `uniform`. Parameters ---------- low : float or array_like of floats, optional Lower boundary of the output interval. All values generated will be greater than or equal to low. The default value is 0. high : float or array_like of floats Upper boundary of the output interval. All values generated will be less than high. The default value is 1.0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``low`` and ``high`` are both scalars. Otherwise, ``np.broadcast(low, high).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized uniform distribution. See Also -------- integers : Discrete uniform distribution, yielding integers. random : Floats uniformly distributed over ``[0, 1)``. Notes ----- The probability density function of the uniform distribution is .. math:: p(x) = \frac{1}{b - a} anywhere within the interval ``[a, b)``, and zero elsewhere. When ``high`` == ``low``, values of ``low`` will be returned. If ``high`` < ``low``, the results are officially undefined and may eventually raise an error, i.e. do not rely on this function to behave when passed arguments satisfying that inequality condition. Examples -------- Draw samples from the distribution: >>> s = np.random.default_rng().uniform(-1,0,1000) All values are within the given interval: >>> np.all(s >= -1) True >>> np.all(s < 0) True Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 15, density=True) >>> plt.plot(bins, np.ones_like(bins), linewidth=2, color='r') >>> plt.show() triangular(left, mode, right, size=None) Draw samples from the triangular distribution over the interval ``[left, right]``. The triangular distribution is a continuous probability distribution with lower limit left, peak at mode, and upper limit right. Unlike the other distributions, these parameters directly define the shape of the pdf. Parameters ---------- left : float or array_like of floats Lower limit. mode : float or array_like of floats The value where the peak of the distribution occurs. The value must fulfill the condition ``left <= mode <= right``. right : float or array_like of floats Upper limit, must be larger than `left`. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``left``, ``mode``, and ``right`` are all scalars. Otherwise, ``np.broadcast(left, mode, right).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized triangular distribution. Notes ----- The probability density function for the triangular distribution is .. math:: P(x;l, m, r) = \begin{cases} \frac{2(x-l)}{(r-l)(m-l)}& \text{for $l \leq x \leq m$},\\ \frac{2(r-x)}{(r-l)(r-m)}& \text{for $m \leq x \leq r$},\\ 0& \text{otherwise}. \end{cases} The triangular distribution is often used in ill-defined problems where the underlying distribution is not known, but some knowledge of the limits and mode exists. Often it is used in simulations. References ---------- .. [1] Wikipedia, "Triangular distribution" https://en.wikipedia.org/wiki/Triangular_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> h = plt.hist(np.random.default_rng().triangular(-3, 0, 8, 100000), bins=200, ... density=True) >>> plt.show() standard_t(df, size=None) Draw samples from a standard Student's t distribution with `df` degrees of freedom. A special case of the hyperbolic distribution. As `df` gets large, the result resembles that of the standard normal distribution (`standard_normal`). Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard Student's t distribution. Notes ----- The probability density function for the t distribution is .. math:: P(x, df) = \frac{\Gamma(\frac{df+1}{2})}{\sqrt{\pi df} \Gamma(\frac{df}{2})}\Bigl( 1+\frac{x^2}{df} \Bigr)^{-(df+1)/2} The t test is based on an assumption that the data come from a Normal distribution. The t test provides a way to test whether the sample mean (that is the mean calculated from the data) is a good estimate of the true mean. The derivation of the t-distribution was first published in 1908 by William Gosset while working for the Guinness Brewery in Dublin. Due to proprietary issues, he had to publish under a pseudonym, and so he used the name Student. References ---------- .. [1] Dalgaard, Peter, "Introductory Statistics With R", Springer, 2002. .. [2] Wikipedia, "Student's t-distribution" https://en.wikipedia.org/wiki/Student's_t-distribution Examples -------- From Dalgaard page 83 [1]_, suppose the daily energy intake for 11 women in kilojoules (kJ) is: >>> intake = np.array([5260., 5470, 5640, 6180, 6390, 6515, 6805, 7515, \ ... 7515, 8230, 8770]) Does their energy intake deviate systematically from the recommended value of 7725 kJ? We have 10 degrees of freedom, so is the sample mean within 95% of the recommended value? >>> s = np.random.default_rng().standard_t(10, size=100000) >>> np.mean(intake) 6753.636363636364 >>> intake.std(ddof=1) 1142.1232221373727 Calculate the t statistic, setting the ddof parameter to the unbiased value so the divisor in the standard deviation will be degrees of freedom, N-1. >>> t = (np.mean(intake)-7725)/(intake.std(ddof=1)/np.sqrt(len(intake))) >>> import matplotlib.pyplot as plt >>> h = plt.hist(s, bins=100, density=True) For a one-sided t-test, how far out in the distribution does the t statistic appear? >>> np.sum(s>> rng = np.random.default_rng() >>> rng.standard_normal() 2.1923875335537315 #random >>> s = rng.standard_normal(8000) >>> s array([ 0.6888893 , 0.78096262, -0.89086505, ..., 0.49876311, # random -0.38672696, -0.4685006 ]) # random >>> s.shape (8000,) >>> s = rng.standard_normal(size=(3, 4, 2)) >>> s.shape (3, 4, 2) Two-by-four array of samples from :math:`N(3, 6.25)`: >>> 3 + 2.5 * rng.standard_normal(size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random shuffle(x, axis=0) Modify a sequence in-place by shuffling its contents. The order of sub-arrays is changed but their contents remains the same. Parameters ---------- x : array_like The array or list to be shuffled. axis : int, optional The axis which `x` is shuffled along. Default is 0. It is only supported on `ndarray` objects. Returns ------- None Examples -------- >>> rng = np.random.default_rng() >>> arr = np.arange(10) >>> rng.shuffle(arr) >>> arr [1 7 5 2 9 4 3 6 0 8] # random >>> arr = np.arange(9).reshape((3, 3)) >>> rng.shuffle(arr) >>> arr array([[3, 4, 5], # random [6, 7, 8], [0, 1, 2]]) >>> arr = np.arange(9).reshape((3, 3)) >>> rng.shuffle(arr, axis=1) >>> arr array([[2, 0, 1], # random [5, 3, 4], [8, 6, 7]]) rayleigh(scale=1.0, size=None) Draw samples from a Rayleigh distribution. The :math:`\chi` and Weibull distributions are generalizations of the Rayleigh. Parameters ---------- scale : float or array_like of floats, optional Scale, also equals the mode. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Rayleigh distribution. Notes ----- The probability density function for the Rayleigh distribution is .. math:: P(x;scale) = \frac{x}{scale^2}e^{\frac{-x^2}{2 \cdotp scale^2}} The Rayleigh distribution would arise, for example, if the East and North components of the wind velocity had identical zero-mean Gaussian distributions. Then the wind speed would have a Rayleigh distribution. References ---------- .. [1] Brighton Webs Ltd., "Rayleigh Distribution," https://web.archive.org/web/20090514091424/http://brighton-webs.co.uk:80/distributions/rayleigh.asp .. [2] Wikipedia, "Rayleigh distribution" https://en.wikipedia.org/wiki/Rayleigh_distribution Examples -------- Draw values from the distribution and plot the histogram >>> from matplotlib.pyplot import hist >>> rng = np.random.default_rng() >>> values = hist(rng.rayleigh(3, 100000), bins=200, density=True) Wave heights tend to follow a Rayleigh distribution. If the mean wave height is 1 meter, what fraction of waves are likely to be larger than 3 meters? >>> meanvalue = 1 >>> modevalue = np.sqrt(2 / np.pi) * meanvalue >>> s = rng.rayleigh(modevalue, 1000000) The percentage of waves larger than 3 meters is: >>> 100.*sum(s>3)/1000000. 0.087300000000000003 # random random(size=None, dtype='d', out=None) Return random floats in the half-open interval [0.0, 1.0). Results are from the "continuous uniform" distribution over the stated interval. To sample :math:`Unif[a, b), b > a` multiply the output of `random` by `(b-a)` and add `a`:: (b - a) * random() + a Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : {str, dtype}, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray of floats Array of random floats of shape `size` (unless ``size=None``, in which case a single float is returned). Examples -------- >>> rng = np.random.default_rng() >>> rng.random() 0.47108547995356098 # random >>> type(rng.random()) >>> rng.random((5,)) array([ 0.30220482, 0.86820401, 0.1654503 , 0.11659149, 0.54323428]) # random Three-by-two array of random numbers from [-5, 0): >>> 5 * rng.random((3, 2)) - 5 array([[-3.99149989, -0.52338984], # random [-2.99091858, -0.79479508], [-1.23204345, -1.75224494]]) power(a, size=None) Draws samples in [0, 1] from a power distribution with positive exponent a - 1. Also known as the power function distribution. Parameters ---------- a : float or array_like of floats Parameter of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized power distribution. Raises ------ ValueError If a < 1. Notes ----- The probability density function is .. math:: P(x; a) = ax^{a-1}, 0 \le x \le 1, a>0. The power function distribution is just the inverse of the Pareto distribution. It may also be seen as a special case of the Beta distribution. It is used, for example, in modeling the over-reporting of insurance claims. References ---------- .. [1] Christian Kleiber, Samuel Kotz, "Statistical size distributions in economics and actuarial sciences", Wiley, 2003. .. [2] Heckert, N. A. and Filliben, James J. "NIST Handbook 148: Dataplot Reference Manual, Volume 2: Let Subcommands and Library Functions", National Institute of Standards and Technology Handbook Series, June 2003. https://www.itl.nist.gov/div898/software/dataplot/refman2/auxillar/powpdf.pdf Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> samples = 1000 >>> s = rng.power(a, samples) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, bins=30) >>> x = np.linspace(0, 1, 100) >>> y = a*x**(a-1.) >>> normed_y = samples*np.diff(bins)[0]*y >>> plt.plot(x, normed_y) >>> plt.show() Compare the power function distribution to the inverse of the Pareto. >>> from scipy import stats # doctest: +SKIP >>> rvs = rng.power(5, 1000000) >>> rvsp = rng.pareto(5, 1000000) >>> xx = np.linspace(0,1,100) >>> powpdf = stats.powerlaw.pdf(xx,5) # doctest: +SKIP >>> plt.figure() >>> plt.hist(rvs, bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('power(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of 1 + Generator.pareto(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of stats.pareto(5)') permutation(x, axis=0) Randomly permute a sequence, or return a permuted range. Parameters ---------- x : int or array_like If `x` is an integer, randomly permute ``np.arange(x)``. If `x` is an array, make a copy and shuffle the elements randomly. axis : int, optional The axis which `x` is shuffled along. Default is 0. Returns ------- out : ndarray Permuted sequence or array range. Examples -------- >>> rng = np.random.default_rng() >>> rng.permutation(10) array([1, 7, 4, 3, 0, 9, 2, 5, 8, 6]) # random >>> rng.permutation([1, 4, 9, 12, 15]) array([15, 1, 9, 4, 12]) # random >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr) array([[6, 7, 8], # random [0, 1, 2], [3, 4, 5]]) >>> rng.permutation("abc") Traceback (most recent call last): ... numpy.AxisError: x must be an integer or at least 1-dimensional >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr, axis=1) array([[0, 2, 1], # random [3, 5, 4], [6, 8, 7]]) pareto(a, size=None) Draw samples from a Pareto II or Lomax distribution with specified shape. The Lomax or Pareto II distribution is a shifted Pareto distribution. The classical Pareto distribution can be obtained from the Lomax distribution by adding 1 and multiplying by the scale parameter ``m`` (see Notes). The smallest value of the Lomax distribution is zero while for the classical Pareto distribution it is ``mu``, where the standard Pareto distribution has location ``mu = 1``. Lomax can also be considered as a simplified version of the Generalized Pareto distribution (available in SciPy), with the scale set to one and the location set to zero. The Pareto distribution must be greater than zero, and is unbounded above. It is also known as the "80-20 rule". In this distribution, 80 percent of the weights are in the lowest 20 percent of the range, while the other 20 percent fill the remaining 80 percent of the range. Parameters ---------- a : float or array_like of floats Shape of the distribution. Must be positive. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Pareto distribution. See Also -------- scipy.stats.lomax : probability density function, distribution or cumulative density function, etc. scipy.stats.genpareto : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Pareto distribution is .. math:: p(x) = \frac{am^a}{x^{a+1}} where :math:`a` is the shape and :math:`m` the scale. The Pareto distribution, named after the Italian economist Vilfredo Pareto, is a power law probability distribution useful in many real world problems. Outside the field of economics it is generally referred to as the Bradford distribution. Pareto developed the distribution to describe the distribution of wealth in an economy. It has also found use in insurance, web page access statistics, oil field sizes, and many other problems, including the download frequency for projects in Sourceforge [1]_. It is one of the so-called "fat-tailed" distributions. References ---------- .. [1] Francis Hunt and Paul Johnson, On the Pareto Distribution of Sourceforge projects. .. [2] Pareto, V. (1896). Course of Political Economy. Lausanne. .. [3] Reiss, R.D., Thomas, M.(2001), Statistical Analysis of Extreme Values, Birkhauser Verlag, Basel, pp 23-30. .. [4] Wikipedia, "Pareto distribution", https://en.wikipedia.org/wiki/Pareto_distribution Examples -------- Draw samples from the distribution: >>> a, m = 3., 2. # shape and mode >>> s = (np.random.default_rng().pareto(a, 1000) + 1) * m Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 100, density=True) >>> fit = a*m**a / bins**(a+1) >>> plt.plot(bins, max(count)*fit/max(fit), linewidth=2, color='r') >>> plt.show() numpy.core.multiarray failed to import normal(loc=0.0, scale=1.0, size=None) Draw random samples from a normal (Gaussian) distribution. The probability density function of the normal distribution, first derived by De Moivre and 200 years later by both Gauss and Laplace independently [2]_, is often called the bell curve because of its characteristic shape (see the example below). The normal distributions occurs often in nature. For example, it describes the commonly occurring distribution of samples influenced by a large number of tiny, random disturbances, each with its own unique distribution [2]_. Parameters ---------- loc : float or array_like of floats Mean ("centre") of the distribution. scale : float or array_like of floats Standard deviation (spread or "width") of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized normal distribution. See Also -------- scipy.stats.norm : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gaussian distribution is .. math:: p(x) = \frac{1}{\sqrt{ 2 \pi \sigma^2 }} e^{ - \frac{ (x - \mu)^2 } {2 \sigma^2} }, where :math:`\mu` is the mean and :math:`\sigma` the standard deviation. The square of the standard deviation, :math:`\sigma^2`, is called the variance. The function has its peak at the mean, and its "spread" increases with the standard deviation (the function reaches 0.607 times its maximum at :math:`x + \sigma` and :math:`x - \sigma` [2]_). This implies that :meth:`normal` is more likely to return samples lying close to the mean, rather than those far away. References ---------- .. [1] Wikipedia, "Normal distribution", https://en.wikipedia.org/wiki/Normal_distribution .. [2] P. R. Peebles Jr., "Central Limit Theorem" in "Probability, Random Variables and Random Signal Principles", 4th ed., 2001, pp. 51, 51, 125. Examples -------- Draw samples from the distribution: >>> mu, sigma = 0, 0.1 # mean and standard deviation >>> s = np.random.default_rng().normal(mu, sigma, 1000) Verify the mean and the variance: >>> abs(mu - np.mean(s)) 0.0 # may vary >>> abs(sigma - np.std(s, ddof=1)) 0.1 # may vary Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 30, density=True) >>> plt.plot(bins, 1/(sigma * np.sqrt(2 * np.pi)) * ... np.exp( - (bins - mu)**2 / (2 * sigma**2) ), ... linewidth=2, color='r') >>> plt.show() Two-by-four array of samples from N(3, 6.25): >>> np.random.default_rng().normal(3, 2.5, size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random noncentral_f(dfnum, dfden, nonc, size=None) Draw samples from the noncentral F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters > 1. `nonc` is the non-centrality parameter. Parameters ---------- dfnum : float or array_like of floats Numerator degrees of freedom, must be > 0. .. versionchanged:: 1.14.0 Earlier NumPy versions required dfnum > 1. dfden : float or array_like of floats Denominator degrees of freedom, must be > 0. nonc : float or array_like of floats Non-centrality parameter, the sum of the squares of the numerator means, must be >= 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum``, ``dfden``, and ``nonc`` are all scalars. Otherwise, ``np.broadcast(dfnum, dfden, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral Fisher distribution. Notes ----- When calculating the power of an experiment (power = probability of rejecting the null hypothesis when a specific alternative is true) the non-central F statistic becomes important. When the null hypothesis is true, the F statistic follows a central F distribution. When the null hypothesis is not true, then it follows a non-central F statistic. References ---------- .. [1] Weisstein, Eric W. "Noncentral F-Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/NoncentralF-Distribution.html .. [2] Wikipedia, "Noncentral F-distribution", https://en.wikipedia.org/wiki/Noncentral_F-distribution Examples -------- In a study, testing for a specific alternative to the null hypothesis requires use of the Noncentral F distribution. We need to calculate the area in the tail of the distribution that exceeds the value of the F distribution for the null hypothesis. We'll plot the two probability distributions for comparison. >>> rng = np.random.default_rng() >>> dfnum = 3 # between group deg of freedom >>> dfden = 20 # within groups degrees of freedom >>> nonc = 3.0 >>> nc_vals = rng.noncentral_f(dfnum, dfden, nonc, 1000000) >>> NF = np.histogram(nc_vals, bins=50, density=True) >>> c_vals = rng.f(dfnum, dfden, 1000000) >>> F = np.histogram(c_vals, bins=50, density=True) >>> import matplotlib.pyplot as plt >>> plt.plot(F[1][1:], F[0]) >>> plt.plot(NF[1][1:], NF[0]) >>> plt.show() noncentral_chisquare(df, nonc, size=None) Draw samples from a noncentral chi-square distribution. The noncentral :math:`\chi^2` distribution is a generalization of the :math:`\chi^2` distribution. Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. .. versionchanged:: 1.10.0 Earlier NumPy versions required dfnum > 1. nonc : float or array_like of floats Non-centrality, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` and ``nonc`` are both scalars. Otherwise, ``np.broadcast(df, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral chi-square distribution. Notes ----- The probability density function for the noncentral Chi-square distribution is .. math:: P(x;df,nonc) = \sum^{\infty}_{i=0} \frac{e^{-nonc/2}(nonc/2)^{i}}{i!} P_{Y_{df+2i}}(x), where :math:`Y_{q}` is the Chi-square with q degrees of freedom. References ---------- .. [1] Wikipedia, "Noncentral chi-squared distribution" https://en.wikipedia.org/wiki/Noncentral_chi-squared_distribution Examples -------- Draw values from the distribution and plot the histogram >>> rng = np.random.default_rng() >>> import matplotlib.pyplot as plt >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() Draw values from a noncentral chisquare with very small noncentrality, and compare to a chisquare. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, .0000001, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> values2 = plt.hist(rng.chisquare(3, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> plt.plot(values[1][0:-1], values[0]-values2[0], 'ob') >>> plt.show() Demonstrate how large values of non-centrality lead to a more symmetric distribution. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() negative_binomial(n, p, size=None) Draw samples from a negative binomial distribution. Samples are drawn from a negative binomial distribution with specified parameters, `n` successes and `p` probability of success where `n` is > 0 and `p` is in the interval [0, 1]. Parameters ---------- n : float or array_like of floats Parameter of the distribution, > 0. p : float or array_like of floats Parameter of the distribution, >= 0 and <=1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``n`` and ``p`` are both scalars. Otherwise, ``np.broadcast(n, p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized negative binomial distribution, where each sample is equal to N, the number of failures that occurred before a total of n successes was reached. Notes ----- The probability mass function of the negative binomial distribution is .. math:: P(N;n,p) = \frac{\Gamma(N+n)}{N!\Gamma(n)}p^{n}(1-p)^{N}, where :math:`n` is the number of successes, :math:`p` is the probability of success, :math:`N+n` is the number of trials, and :math:`\Gamma` is the gamma function. When :math:`n` is an integer, :math:`\frac{\Gamma(N+n)}{N!\Gamma(n)} = \binom{N+n-1}{N}`, which is the more common form of this term in the the pmf. The negative binomial distribution gives the probability of N failures given n successes, with a success on the last trial. If one throws a die repeatedly until the third time a "1" appears, then the probability distribution of the number of non-"1"s that appear before the third "1" is a negative binomial distribution. References ---------- .. [1] Weisstein, Eric W. "Negative Binomial Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/NegativeBinomialDistribution.html .. [2] Wikipedia, "Negative binomial distribution", https://en.wikipedia.org/wiki/Negative_binomial_distribution Examples -------- Draw samples from the distribution: A real world example. A company drills wild-cat oil exploration wells, each with an estimated probability of success of 0.1. What is the probability of having one success for each successive well, that is what is the probability of a single success after drilling 5 wells, after 6 wells, etc.? >>> s = np.random.default_rng().negative_binomial(1, 0.1, 100000) >>> for i in range(1, 11): # doctest: +SKIP ... probability = sum(s>> mean = [0, 0] >>> cov = [[1, 0], [0, 100]] # diagonal covariance Diagonal covariance means that points are oriented along x or y-axis: >>> import matplotlib.pyplot as plt >>> x, y = np.random.default_rng().multivariate_normal(mean, cov, 5000).T >>> plt.plot(x, y, 'x') >>> plt.axis('equal') >>> plt.show() Note that the covariance matrix must be positive semidefinite (a.k.a. nonnegative-definite). Otherwise, the behavior of this method is undefined and backwards compatibility is not guaranteed. References ---------- .. [1] Papoulis, A., "Probability, Random Variables, and Stochastic Processes," 3rd ed., New York: McGraw-Hill, 1991. .. [2] Duda, R. O., Hart, P. E., and Stork, D. G., "Pattern Classification," 2nd ed., New York: Wiley, 2001. Examples -------- >>> mean = (1, 2) >>> cov = [[1, 0], [0, 1]] >>> rng = np.random.default_rng() >>> x = rng.multivariate_normal(mean, cov, (3, 3)) >>> x.shape (3, 3, 2) We can use a different method other than the default to factorize cov: >>> y = rng.multivariate_normal(mean, cov, (3, 3), method='cholesky') >>> y.shape (3, 3, 2) The following is probably true, given that 0.6 is roughly twice the standard deviation: >>> list((x[0,0,:] - mean) < 0.6) [True, True] # random multivariate_hypergeometric(colors, nsample, size=None, method='marginals') Generate variates from a multivariate hypergeometric distribution. The multivariate hypergeometric distribution is a generalization of the hypergeometric distribution. Choose ``nsample`` items at random without replacement from a collection with ``N`` distinct types. ``N`` is the length of ``colors``, and the values in ``colors`` are the number of occurrences of that type in the collection. The total number of items in the collection is ``sum(colors)``. Each random variate generated by this function is a vector of length ``N`` holding the counts of the different types that occurred in the ``nsample`` items. The name ``colors`` comes from a common description of the distribution: it is the probability distribution of the number of marbles of each color selected without replacement from an urn containing marbles of different colors; ``colors[i]`` is the number of marbles in the urn with color ``i``. Parameters ---------- colors : sequence of integers The number of each type of item in the collection from which a sample is drawn. The values in ``colors`` must be nonnegative. To avoid loss of precision in the algorithm, ``sum(colors)`` must be less than ``10**9`` when `method` is "marginals". nsample : int The number of items selected. ``nsample`` must not be greater than ``sum(colors)``. size : int or tuple of ints, optional The number of variates to generate, either an integer or a tuple holding the shape of the array of variates. If the given size is, e.g., ``(k, m)``, then ``k * m`` variates are drawn, where one variate is a vector of length ``len(colors)``, and the return value has shape ``(k, m, len(colors))``. If `size` is an integer, the output has shape ``(size, len(colors))``. Default is None, in which case a single variate is returned as an array with shape ``(len(colors),)``. method : string, optional Specify the algorithm that is used to generate the variates. Must be 'count' or 'marginals' (the default). See the Notes for a description of the methods. Returns ------- variates : ndarray Array of variates drawn from the multivariate hypergeometric distribution. See Also -------- hypergeometric : Draw samples from the (univariate) hypergeometric distribution. Notes ----- The two methods do not return the same sequence of variates. The "count" algorithm is roughly equivalent to the following numpy code:: choices = np.repeat(np.arange(len(colors)), colors) selection = np.random.choice(choices, nsample, replace=False) variate = np.bincount(selection, minlength=len(colors)) The "count" algorithm uses a temporary array of integers with length ``sum(colors)``. The "marginals" algorithm generates a variate by using repeated calls to the univariate hypergeometric sampler. It is roughly equivalent to:: variate = np.zeros(len(colors), dtype=np.int64) # `remaining` is the cumulative sum of `colors` from the last # element to the first; e.g. if `colors` is [3, 1, 5], then # `remaining` is [9, 6, 5]. remaining = np.cumsum(colors[::-1])[::-1] for i in range(len(colors)-1): if nsample < 1: break variate[i] = hypergeometric(colors[i], remaining[i+1], nsample) nsample -= variate[i] variate[-1] = nsample The default method is "marginals". For some cases (e.g. when `colors` contains relatively small integers), the "count" method can be significantly faster than the "marginals" method. If performance of the algorithm is important, test the two methods with typical inputs to decide which works best. .. versionadded:: 1.18.0 Examples -------- >>> colors = [16, 8, 4] >>> seed = 4861946401452 >>> gen = np.random.Generator(np.random.PCG64(seed)) >>> gen.multivariate_hypergeometric(colors, 6) array([5, 0, 1]) >>> gen.multivariate_hypergeometric(colors, 6, size=3) array([[5, 0, 1], [2, 2, 2], [3, 3, 0]]) >>> gen.multivariate_hypergeometric(colors, 6, size=(2, 2)) array([[[3, 2, 1], [3, 2, 1]], [[4, 1, 1], [3, 2, 1]]]) multinomial(n, pvals, size=None) Draw samples from a multinomial distribution. The multinomial distribution is a multivariate generalization of the binomial distribution. Take an experiment with one of ``p`` possible outcomes. An example of such an experiment is throwing a dice, where the outcome can be 1 through 6. Each sample drawn from the distribution represents `n` such experiments. Its values, ``X_i = [X_0, X_1, ..., X_p]``, represent the number of times the outcome was ``i``. Parameters ---------- n : int or array-like of ints Number of experiments. pvals : sequence of floats, length p Probabilities of each of the ``p`` different outcomes. These must sum to 1 (however, the last element is always assumed to account for the remaining probability, as long as ``sum(pvals[:-1]) <= 1)``. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- out : ndarray The drawn samples, of shape *size*, if that was provided. If not, the shape is ``(N,)``. In other words, each entry ``out[i,j,...,:]`` is an N-dimensional value drawn from the distribution. Examples -------- Throw a dice 20 times: >>> rng = np.random.default_rng() >>> rng.multinomial(20, [1/6.]*6, size=1) array([[4, 1, 7, 5, 2, 1]]) # random It landed 4 times on 1, once on 2, etc. Now, throw the dice 20 times, and 20 times again: >>> rng.multinomial(20, [1/6.]*6, size=2) array([[3, 4, 3, 3, 4, 3], [2, 4, 3, 4, 0, 7]]) # random For the first run, we threw 3 times 1, 4 times 2, etc. For the second, we threw 2 times 1, 4 times 2, etc. Now, do one experiment throwing the dice 10 time, and 10 times again, and another throwing the dice 20 times, and 20 times again: >>> rng.multinomial([[10], [20]], [1/6.]*6, size=2) array([[[2, 4, 0, 1, 2, 1], [1, 3, 0, 3, 1, 2]], [[1, 4, 4, 4, 4, 3], [3, 3, 2, 5, 5, 2]]]) # random The first array shows the outcomes of throwing the dice 10 times, and the second shows the outcomes from throwing the dice 20 times. A loaded die is more likely to land on number 6: >>> rng.multinomial(100, [1/7.]*5 + [2/7.]) array([11, 16, 14, 17, 16, 26]) # random The probability inputs should be normalized. As an implementation detail, the value of the last entry is ignored and assumed to take up any leftover probability mass, but this should not be relied on. A biased coin which has twice as much weight on one side as on the other should be sampled like so: >>> rng.multinomial(100, [1.0 / 3, 2.0 / 3]) # RIGHT array([38, 62]) # random not like: >>> rng.multinomial(100, [1.0, 2.0]) # WRONG Traceback (most recent call last): ValueError: pvals < 0, pvals > 1 or pvals contains NaNs logseries(p, size=None) Draw samples from a logarithmic series distribution. Samples are drawn from a log series distribution with specified shape parameter, 0 < ``p`` < 1. Parameters ---------- p : float or array_like of floats Shape parameter for the distribution. Must be in the range (0, 1). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logarithmic series distribution. See Also -------- scipy.stats.logser : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function for the Log Series distribution is .. math:: P(k) = \frac{-p^k}{k \ln(1-p)}, where p = probability. The log series distribution is frequently used to represent species richness and occurrence, first proposed by Fisher, Corbet, and Williams in 1943 [2]. It may also be used to model the numbers of occupants seen in cars [3]. References ---------- .. [1] Buzas, Martin A.; Culver, Stephen J., Understanding regional species diversity through the log series distribution of occurrences: BIODIVERSITY RESEARCH Diversity & Distributions, Volume 5, Number 5, September 1999 , pp. 187-195(9). .. [2] Fisher, R.A,, A.S. Corbet, and C.B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology, 12:42-58. .. [3] D. J. Hand, F. Daly, D. Lunn, E. Ostrowski, A Handbook of Small Data Sets, CRC Press, 1994. .. [4] Wikipedia, "Logarithmic distribution", https://en.wikipedia.org/wiki/Logarithmic_distribution Examples -------- Draw samples from the distribution: >>> a = .6 >>> s = np.random.default_rng().logseries(a, 10000) >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s) # plot against distribution >>> def logseries(k, p): ... return -p**k/(k*np.log(1-p)) >>> plt.plot(bins, logseries(bins, a) * count.max()/ ... logseries(bins, a).max(), 'r') >>> plt.show() lognormal(mean=0.0, sigma=1.0, size=None) Draw samples from a log-normal distribution. Draw samples from a log-normal distribution with specified mean, standard deviation, and array shape. Note that the mean and standard deviation are not the values for the distribution itself, but of the underlying normal distribution it is derived from. Parameters ---------- mean : float or array_like of floats, optional Mean value of the underlying normal distribution. Default is 0. sigma : float or array_like of floats, optional Standard deviation of the underlying normal distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``sigma`` are both scalars. Otherwise, ``np.broadcast(mean, sigma).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized log-normal distribution. See Also -------- scipy.stats.lognorm : probability density function, distribution, cumulative density function, etc. Notes ----- A variable `x` has a log-normal distribution if `log(x)` is normally distributed. The probability density function for the log-normal distribution is: .. math:: p(x) = \frac{1}{\sigma x \sqrt{2\pi}} e^{(-\frac{(ln(x)-\mu)^2}{2\sigma^2})} where :math:`\mu` is the mean and :math:`\sigma` is the standard deviation of the normally distributed logarithm of the variable. A log-normal distribution results if a random variable is the *product* of a large number of independent, identically-distributed variables in the same way that a normal distribution results if the variable is the *sum* of a large number of independent, identically-distributed variables. References ---------- .. [1] Limpert, E., Stahel, W. A., and Abbt, M., "Log-normal Distributions across the Sciences: Keys and Clues," BioScience, Vol. 51, No. 5, May, 2001. https://stat.ethz.ch/~stahel/lognormal/bioscience.pdf .. [2] Reiss, R.D. and Thomas, M., "Statistical Analysis of Extreme Values," Basel: Birkhauser Verlag, 2001, pp. 31-32. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, sigma = 3., 1. # mean and standard deviation >>> s = rng.lognormal(mu, sigma, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 100, density=True, align='mid') >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, linewidth=2, color='r') >>> plt.axis('tight') >>> plt.show() Demonstrate that taking the products of random samples from a uniform distribution can be fit well by a log-normal probability density function. >>> # Generate a thousand samples: each is the product of 100 random >>> # values, drawn from a normal distribution. >>> rng = rng >>> b = [] >>> for i in range(1000): ... a = 10. + rng.standard_normal(100) ... b.append(np.product(a)) >>> b = np.array(b) / np.min(b) # scale values to be positive >>> count, bins, ignored = plt.hist(b, 100, density=True, align='mid') >>> sigma = np.std(np.log(b)) >>> mu = np.mean(np.log(b)) >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, color='r', linewidth=2) >>> plt.show() logistic(loc=0.0, scale=1.0, size=None) Draw samples from a logistic distribution. Samples are drawn from a logistic distribution with specified parameters, loc (location or mean, also median), and scale (>0). Parameters ---------- loc : float or array_like of floats, optional Parameter of the distribution. Default is 0. scale : float or array_like of floats, optional Parameter of the distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logistic distribution. See Also -------- scipy.stats.logistic : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Logistic distribution is .. math:: P(x) = P(x) = \frac{e^{-(x-\mu)/s}}{s(1+e^{-(x-\mu)/s})^2}, where :math:`\mu` = location and :math:`s` = scale. The Logistic distribution is used in Extreme Value problems where it can act as a mixture of Gumbel distributions, in Epidemiology, and by the World Chess Federation (FIDE) where it is used in the Elo ranking system, assuming the performance of each player is a logistically distributed random variable. References ---------- .. [1] Reiss, R.-D. and Thomas M. (2001), "Statistical Analysis of Extreme Values, from Insurance, Finance, Hydrology and Other Fields," Birkhauser Verlag, Basel, pp 132-133. .. [2] Weisstein, Eric W. "Logistic Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/LogisticDistribution.html .. [3] Wikipedia, "Logistic-distribution", https://en.wikipedia.org/wiki/Logistic_distribution Examples -------- Draw samples from the distribution: >>> loc, scale = 10, 1 >>> s = np.random.default_rng().logistic(loc, scale, 10000) >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, bins=50) # plot against distribution >>> def logist(x, loc, scale): ... return np.exp((loc-x)/scale)/(scale*(1+np.exp((loc-x)/scale))**2) >>> lgst_val = logist(bins, loc, scale) >>> plt.plot(bins, lgst_val * count.max() / lgst_val.max()) >>> plt.show() integers(low, high=None, size=None, dtype='int64', endpoint=False) Return random integers from `low` (inclusive) to `high` (exclusive), or if endpoint=True, `low` (inclusive) to `high` (inclusive). Replaces `RandomState.randint` (with endpoint=False) and `RandomState.random_integers` (with endpoint=True) Return random integers from the "discrete uniform" distribution of the specified dtype. If `high` is None (the default), then results are from 0 to `low`. Parameters ---------- low : int or array-like of ints Lowest (signed) integers to be drawn from the distribution (unless ``high=None``, in which case this parameter is 0 and this value is used for `high`). high : int or array-like of ints, optional If provided, one above the largest (signed) integer to be drawn from the distribution (see above for behavior if ``high=None``). If array-like, must contain integer values size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : {str, dtype}, optional Desired dtype of the result. All dtypes are determined by their name, i.e., 'int64', 'int', etc, so byteorder is not available and a specific precision may have different C types depending on the platform. The default value is `np.int_`. endpoint : bool, optional If true, sample from the interval [low, high] instead of the default [low, high) Defaults to False Returns ------- out : int or ndarray of ints `size`-shaped array of random integers from the appropriate distribution, or a single such random int if `size` not provided. Notes ----- When using broadcasting with uint64 dtypes, the maximum value (2**64) cannot be represented as a standard integer type. The high array (or low if high is None) must have object dtype, e.g., array([2**64]). Examples -------- >>> rng = np.random.default_rng() >>> rng.integers(2, size=10) array([1, 0, 0, 0, 1, 1, 0, 0, 1, 0]) # random >>> rng.integers(1, size=10) array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0]) Generate a 2 x 4 array of ints between 0 and 4, inclusive: >>> rng.integers(5, size=(2, 4)) array([[4, 0, 2, 1], [3, 2, 2, 0]]) # random Generate a 1 x 3 array with 3 different upper bounds >>> rng.integers(1, [3, 5, 10]) array([2, 2, 9]) # random Generate a 1 by 3 array with 3 different lower bounds >>> rng.integers([1, 5, 7], 10) array([9, 8, 7]) # random Generate a 2 by 4 array using broadcasting with dtype of uint8 >>> rng.integers([1, 3, 5, 7], [[10], [20]], dtype=np.uint8) array([[ 8, 6, 9, 7], [ 1, 16, 9, 12]], dtype=uint8) # random References ---------- .. [1] Daniel Lemire., "Fast Random Integer Generation in an Interval", ACM Transactions on Modeling and Computer Simulation 29 (1), 2019, http://arxiv.org/abs/1805.10941. hypergeometric(ngood, nbad, nsample, size=None) Draw samples from a Hypergeometric distribution. Samples are drawn from a hypergeometric distribution with specified parameters, `ngood` (ways to make a good selection), `nbad` (ways to make a bad selection), and `nsample` (number of items sampled, which is less than or equal to the sum ``ngood + nbad``). Parameters ---------- ngood : int or array_like of ints Number of ways to make a good selection. Must be nonnegative and less than 10**9. nbad : int or array_like of ints Number of ways to make a bad selection. Must be nonnegative and less than 10**9. nsample : int or array_like of ints Number of items sampled. Must be nonnegative and less than ``ngood + nbad``. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if `ngood`, `nbad`, and `nsample` are all scalars. Otherwise, ``np.broadcast(ngood, nbad, nsample).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized hypergeometric distribution. Each sample is the number of good items within a randomly selected subset of size `nsample` taken from a set of `ngood` good items and `nbad` bad items. See Also -------- multivariate_hypergeometric : Draw samples from the multivariate hypergeometric distribution. scipy.stats.hypergeom : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Hypergeometric distribution is .. math:: P(x) = \frac{\binom{g}{x}\binom{b}{n-x}}{\binom{g+b}{n}}, where :math:`0 \le x \le n` and :math:`n-b \le x \le g` for P(x) the probability of ``x`` good results in the drawn sample, g = `ngood`, b = `nbad`, and n = `nsample`. Consider an urn with black and white marbles in it, `ngood` of them are black and `nbad` are white. If you draw `nsample` balls without replacement, then the hypergeometric distribution describes the distribution of black balls in the drawn sample. Note that this distribution is very similar to the binomial distribution, except that in this case, samples are drawn without replacement, whereas in the Binomial case samples are drawn with replacement (or the sample space is infinite). As the sample space becomes large, this distribution approaches the binomial. The arguments `ngood` and `nbad` each must be less than `10**9`. For extremely large arguments, the algorithm that is used to compute the samples [4]_ breaks down because of loss of precision in floating point calculations. For such large values, if `nsample` is not also large, the distribution can be approximated with the binomial distribution, `binomial(n=nsample, p=ngood/(ngood + nbad))`. References ---------- .. [1] Lentner, Marvin, "Elementary Applied Statistics", Bogden and Quigley, 1972. .. [2] Weisstein, Eric W. "Hypergeometric Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/HypergeometricDistribution.html .. [3] Wikipedia, "Hypergeometric distribution", https://en.wikipedia.org/wiki/Hypergeometric_distribution .. [4] Stadlober, Ernst, "The ratio of uniforms approach for generating discrete random variates", Journal of Computational and Applied Mathematics, 31, pp. 181-189 (1990). Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> ngood, nbad, nsamp = 100, 2, 10 # number of good, number of bad, and number of samples >>> s = rng.hypergeometric(ngood, nbad, nsamp, 1000) >>> from matplotlib.pyplot import hist >>> hist(s) # note that it is very unlikely to grab both bad items Suppose you have an urn with 15 white and 15 black marbles. If you pull 15 marbles at random, how likely is it that 12 or more of them are one color? >>> s = rng.hypergeometric(15, 15, 15, 100000) >>> sum(s>=12)/100000. + sum(s<=3)/100000. # answer = 0.003 ... pretty unlikely! gumbel(loc=0.0, scale=1.0, size=None) Draw samples from a Gumbel distribution. Draw samples from a Gumbel distribution with specified location and scale. For more information on the Gumbel distribution, see Notes and References below. Parameters ---------- loc : float or array_like of floats, optional The location of the mode of the distribution. Default is 0. scale : float or array_like of floats, optional The scale parameter of the distribution. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Gumbel distribution. See Also -------- scipy.stats.gumbel_l scipy.stats.gumbel_r scipy.stats.genextreme weibull Notes ----- The Gumbel (or Smallest Extreme Value (SEV) or the Smallest Extreme Value Type I) distribution is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. The Gumbel is a special case of the Extreme Value Type I distribution for maximums from distributions with "exponential-like" tails. The probability density for the Gumbel distribution is .. math:: p(x) = \frac{e^{-(x - \mu)/ \beta}}{\beta} e^{ -e^{-(x - \mu)/ \beta}}, where :math:`\mu` is the mode, a location parameter, and :math:`\beta` is the scale parameter. The Gumbel (named for German mathematician Emil Julius Gumbel) was used very early in the hydrology literature, for modeling the occurrence of flood events. It is also used for modeling maximum wind speed and rainfall rates. It is a "fat-tailed" distribution - the probability of an event in the tail of the distribution is larger than if one used a Gaussian, hence the surprisingly frequent occurrence of 100-year floods. Floods were initially modeled as a Gaussian process, which underestimated the frequency of extreme events. It is one of a class of extreme value distributions, the Generalized Extreme Value (GEV) distributions, which also includes the Weibull and Frechet. The function has a mean of :math:`\mu + 0.57721\beta` and a variance of :math:`\frac{\pi^2}{6}\beta^2`. References ---------- .. [1] Gumbel, E. J., "Statistics of Extremes," New York: Columbia University Press, 1958. .. [2] Reiss, R.-D. and Thomas, M., "Statistical Analysis of Extreme Values from Insurance, Finance, Hydrology and Other Fields," Basel: Birkhauser Verlag, 2001. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, beta = 0, 0.1 # location and scale >>> s = rng.gumbel(mu, beta, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 30, density=True) >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp( -np.exp( -(bins - mu) /beta) ), ... linewidth=2, color='r') >>> plt.show() Show how an extreme value distribution can arise from a Gaussian process and compare to a Gaussian: >>> means = [] >>> maxima = [] >>> for i in range(0,1000) : ... a = rng.normal(mu, beta, 1000) ... means.append(a.mean()) ... maxima.append(a.max()) >>> count, bins, ignored = plt.hist(maxima, 30, density=True) >>> beta = np.std(maxima) * np.sqrt(6) / np.pi >>> mu = np.mean(maxima) - 0.57721*beta >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp(-np.exp(-(bins - mu)/beta)), ... linewidth=2, color='r') >>> plt.plot(bins, 1/(beta * np.sqrt(2 * np.pi)) ... * np.exp(-(bins - mu)**2 / (2 * beta**2)), ... linewidth=2, color='g') >>> plt.show() geometric(p, size=None) Draw samples from the geometric distribution. Bernoulli trials are experiments with one of two outcomes: success or failure (an example of such an experiment is flipping a coin). The geometric distribution models the number of trials that must be run in order to achieve success. It is therefore supported on the positive integers, ``k = 1, 2, ...``. The probability mass function of the geometric distribution is .. math:: f(k) = (1 - p)^{k - 1} p where `p` is the probability of success of an individual trial. Parameters ---------- p : float or array_like of floats The probability of success of an individual trial. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized geometric distribution. Examples -------- Draw ten thousand values from the geometric distribution, with the probability of an individual success equal to 0.35: >>> z = np.random.default_rng().geometric(p=0.35, size=10000) How many trials succeeded after a single run? >>> (z == 1).sum() / 10000. 0.34889999999999999 #random gamma(shape, scale=1.0, size=None) Draw samples from a Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, `shape` (sometimes designated "k") and `scale` (sometimes designated "theta"), where both parameters are > 0. Parameters ---------- shape : float or array_like of floats The shape of the gamma distribution. Must be non-negative. scale : float or array_like of floats, optional The scale of the gamma distribution. Must be non-negative. Default is equal to 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(shape, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 2. # mean=4, std=2*sqrt(2) >>> s = np.random.default_rng().gamma(shape, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, ignored = plt.hist(s, 50, density=True) >>> y = bins**(shape-1)*(np.exp(-bins/scale) / # doctest: +SKIP ... (sps.gamma(shape)*scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() f(dfnum, dfden, size=None) Draw samples from an F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters must be greater than zero. The random variate of the F distribution (also known as the Fisher distribution) is a continuous probability distribution that arises in ANOVA tests, and is the ratio of two chi-square variates. Parameters ---------- dfnum : float or array_like of floats Degrees of freedom in numerator, must be > 0. dfden : float or array_like of float Degrees of freedom in denominator, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum`` and ``dfden`` are both scalars. Otherwise, ``np.broadcast(dfnum, dfden).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Fisher distribution. See Also -------- scipy.stats.f : probability density function, distribution or cumulative density function, etc. Notes ----- The F statistic is used to compare in-group variances to between-group variances. Calculating the distribution depends on the sampling, and so it is a function of the respective degrees of freedom in the problem. The variable `dfnum` is the number of samples minus one, the between-groups degrees of freedom, while `dfden` is the within-groups degrees of freedom, the sum of the number of samples in each group minus the number of groups. References ---------- .. [1] Glantz, Stanton A. "Primer of Biostatistics.", McGraw-Hill, Fifth Edition, 2002. .. [2] Wikipedia, "F-distribution", https://en.wikipedia.org/wiki/F-distribution Examples -------- An example from Glantz[1], pp 47-40: Two groups, children of diabetics (25 people) and children from people without diabetes (25 controls). Fasting blood glucose was measured, case group had a mean value of 86.1, controls had a mean value of 82.2. Standard deviations were 2.09 and 2.49 respectively. Are these data consistent with the null hypothesis that the parents diabetic status does not affect their children's blood glucose levels? Calculating the F statistic from the data gives a value of 36.01. Draw samples from the distribution: >>> dfnum = 1. # between group degrees of freedom >>> dfden = 48. # within groups degrees of freedom >>> s = np.random.default_rng().f(dfnum, dfden, 1000) The lower bound for the top 1% of the samples is : >>> np.sort(s)[-10] 7.61988120985 # random So there is about a 1% chance that the F statistic will exceed 7.62, the measured value is 36, so the null hypothesis is rejected at the 1% level. choice(a, size=None, replace=True, p=None, axis=0): Generates a random sample from a given 1-D array Parameters ---------- a : 1-D array-like or int If an ndarray, a random sample is generated from its elements. If an int, the random sample is generated as if a were np.arange(a) size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn from the 1-d `a`. If `a` has more than one dimension, the `size` shape will be inserted into the `axis` dimension, so the output ``ndim`` will be ``a.ndim - 1 + len(size)``. Default is None, in which case a single value is returned. replace : boolean, optional Whether the sample is with or without replacement p : 1-D array-like, optional The probabilities associated with each entry in a. If not given the sample assumes a uniform distribution over all entries in a. axis : int, optional The axis along which the selection is performed. The default, 0, selects by row. shuffle : boolean, optional Whether the sample is shuffled when sampling without replacement. Default is True, False provides a speedup. Returns ------- samples : single item or ndarray The generated random samples Raises ------ ValueError If a is an int and less than zero, if p is not 1-dimensional, if a is array-like with a size 0, if p is not a vector of probabilities, if a and p have different lengths, or if replace=False and the sample size is greater than the population size. See Also -------- integers, shuffle, permutation Examples -------- Generate a uniform random sample from np.arange(5) of size 3: >>> rng = np.random.default_rng() >>> rng.choice(5, 3) array([0, 3, 4]) # random >>> #This is equivalent to rng.integers(0,5,3) Generate a non-uniform random sample from np.arange(5) of size 3: >>> rng.choice(5, 3, p=[0.1, 0, 0.3, 0.6, 0]) array([3, 3, 0]) # random Generate a uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False) array([3,1,0]) # random >>> #This is equivalent to rng.permutation(np.arange(5))[:3] Generate a non-uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False, p=[0.1, 0, 0.3, 0.6, 0]) array([2, 3, 0]) # random Any of the above can be repeated with an arbitrary array-like instead of just integers. For instance: >>> aa_milne_arr = ['pooh', 'rabbit', 'piglet', 'Christopher'] >>> rng.choice(aa_milne_arr, 5, p=[0.5, 0.1, 0.1, 0.3]) array(['pooh', 'pooh', 'pooh', 'Christopher', 'piglet'], # random dtype=' 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized chi-square distribution. Raises ------ ValueError When `df` <= 0 or when an inappropriate `size` (e.g. ``size=-1``) is given. Notes ----- The variable obtained by summing the squares of `df` independent, standard normally distributed random variables: .. math:: Q = \sum_{i=0}^{\mathtt{df}} X^2_i is chi-square distributed, denoted .. math:: Q \sim \chi^2_k. The probability density function of the chi-squared distribution is .. math:: p(x) = \frac{(1/2)^{k/2}}{\Gamma(k/2)} x^{k/2 - 1} e^{-x/2}, where :math:`\Gamma` is the gamma function, .. math:: \Gamma(x) = \int_0^{-\infty} t^{x - 1} e^{-t} dt. References ---------- .. [1] NIST "Engineering Statistics Handbook" https://www.itl.nist.gov/div898/handbook/eda/section3/eda3666.htm Examples -------- >>> np.random.default_rng().chisquare(2,4) array([ 1.89920014, 9.00867716, 3.13710533, 5.62318272]) # random bytes(length) Return random bytes. Parameters ---------- length : int Number of random bytes. Returns ------- out : str String of length `length`. Examples -------- >>> np.random.default_rng().bytes(10) ' eh\x85\x022SZ\xbf\xa4' #random binomial(n, p, size=None) Draw samples from a binomial distribution. Samples are drawn from a binomial distribution with specified parameters, n trials and p probability of success where n an integer >= 0 and p is in the interval [0,1]. (n may be input as a float, but it is truncated to an integer in use) Parameters ---------- n : int or array_like of ints Parameter of the distribution, >= 0. Floats are also accepted, but they will be truncated to integers. p : float or array_like of floats Parameter of the distribution, >= 0 and <=1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``n`` and ``p`` are both scalars. Otherwise, ``np.broadcast(n, p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized binomial distribution, where each sample is equal to the number of successes over the n trials. See Also -------- scipy.stats.binom : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the binomial distribution is .. math:: P(N) = \binom{n}{N}p^N(1-p)^{n-N}, where :math:`n` is the number of trials, :math:`p` is the probability of success, and :math:`N` is the number of successes. When estimating the standard error of a proportion in a population by using a random sample, the normal distribution works well unless the product p*n <=5, where p = population proportion estimate, and n = number of samples, in which case the binomial distribution is used instead. For example, a sample of 15 people shows 4 who are left handed, and 11 who are right handed. Then p = 4/15 = 27%. 0.27*15 = 4, so the binomial distribution should be used in this case. References ---------- .. [1] Dalgaard, Peter, "Introductory Statistics with R", Springer-Verlag, 2002. .. [2] Glantz, Stanton A. "Primer of Biostatistics.", McGraw-Hill, Fifth Edition, 2002. .. [3] Lentner, Marvin, "Elementary Applied Statistics", Bogden and Quigley, 1972. .. [4] Weisstein, Eric W. "Binomial Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/BinomialDistribution.html .. [5] Wikipedia, "Binomial distribution", https://en.wikipedia.org/wiki/Binomial_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> n, p = 10, .5 # number of trials, probability of each trial >>> s = rng.binomial(n, p, 1000) # result of flipping a coin 10 times, tested 1000 times. A real world example. A company drills 9 wild-cat oil exploration wells, each with an estimated probability of success of 0.1. All nine wells fail. What is the probability of that happening? Let's do 20,000 trials of the model, and count the number that generate zero positive results. >>> sum(rng.binomial(9, 0.1, 20000) == 0)/20000. # answer = 0.38885, or 38%. a must be greater than 0 unless no samples are taken'a' cannot be empty unless no samples are takenWhen method is 'count', sum(colors) must not exceed %dGenerator.permutation (line 4308)Generator.multinomial (line 3661)Axis argument is only supported on ndarray objects wald(mean, scale, size=None) Draw samples from a Wald, or inverse Gaussian, distribution. As the scale approaches infinity, the distribution becomes more like a Gaussian. Some references claim that the Wald is an inverse Gaussian with mean equal to 1, but this is by no means universal. The inverse Gaussian distribution was first studied in relationship to Brownian motion. In 1956 M.C.K. Tweedie used the name inverse Gaussian because there is an inverse relationship between the time to cover a unit distance and distance covered in unit time. Parameters ---------- mean : float or array_like of floats Distribution mean, must be > 0. scale : float or array_like of floats Scale parameter, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(mean, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Wald distribution. Notes ----- The probability density function for the Wald distribution is .. math:: P(x;mean,scale) = \sqrt{\frac{scale}{2\pi x^3}}e^ \frac{-scale(x-mean)^2}{2\cdotp mean^2x} As noted above the inverse Gaussian distribution first arise from attempts to model Brownian motion. It is also a competitor to the Weibull for use in reliability modeling and modeling stock returns and interest rate processes. References ---------- .. [1] Brighton Webs Ltd., Wald Distribution, https://web.archive.org/web/20090423014010/http://www.brighton-webs.co.uk:80/distributions/wald.asp .. [2] Chhikara, Raj S., and Folks, J. Leroy, "The Inverse Gaussian Distribution: Theory : Methodology, and Applications", CRC Press, 1988. .. [3] Wikipedia, "Inverse Gaussian distribution" https://en.wikipedia.org/wiki/Inverse_Gaussian_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> h = plt.hist(np.random.default_rng().wald(3, 2, 100000), bins=200, density=True) >>> plt.show() standard_gamma(shape, size=None, dtype='d', out=None) Draw samples from a standard Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, shape (sometimes designated "k") and scale=1. Parameters ---------- shape : float or array_like of floats Parameter, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` is a scalar. Otherwise, ``np.array(shape).size`` samples are drawn. dtype : {str, dtype}, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 1. # mean and width >>> s = np.random.default_rng().standard_gamma(shape, 1000000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, ignored = plt.hist(s, 50, density=True) >>> y = bins**(shape-1) * ((np.exp(-bins/scale))/ # doctest: +SKIP ... (sps.gamma(shape) * scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() standard_exponential(size=None, dtype='d', method='zig', out=None) Draw samples from the standard exponential distribution. `standard_exponential` is identical to the exponential distribution with a scale parameter of 1. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. method : str, optional Either 'inv' or 'zig'. 'inv' uses the default inverse CDF method. 'zig' uses the much faster Ziggurat method of Marsaglia and Tsang. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray Drawn samples. Examples -------- Output a 3x8000 array: >>> n = np.random.default_rng().standard_exponential((3, 8000)) standard_cauchy(size=None) Draw samples from a standard Cauchy distribution with mode = 0. Also known as the Lorentz distribution. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- samples : ndarray or scalar The drawn samples. Notes ----- The probability density function for the full Cauchy distribution is .. math:: P(x; x_0, \gamma) = \frac{1}{\pi \gamma \bigl[ 1+ (\frac{x-x_0}{\gamma})^2 \bigr] } and the Standard Cauchy distribution just sets :math:`x_0=0` and :math:`\gamma=1` The Cauchy distribution arises in the solution to the driven harmonic oscillator problem, and also describes spectral line broadening. It also describes the distribution of values at which a line tilted at a random angle will cut the x axis. When studying hypothesis tests that assume normality, seeing how the tests perform on data from a Cauchy distribution is a good indicator of their sensitivity to a heavy-tailed distribution, since the Cauchy looks very much like a Gaussian distribution, but with heavier tails. References ---------- .. [1] NIST/SEMATECH e-Handbook of Statistical Methods, "Cauchy Distribution", https://www.itl.nist.gov/div898/handbook/eda/section3/eda3663.htm .. [2] Weisstein, Eric W. "Cauchy Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/CauchyDistribution.html .. [3] Wikipedia, "Cauchy distribution" https://en.wikipedia.org/wiki/Cauchy_distribution Examples -------- Draw samples and plot the distribution: >>> import matplotlib.pyplot as plt >>> s = np.random.default_rng().standard_cauchy(1000000) >>> s = s[(s>-25) & (s<25)] # truncate distribution so it plots well >>> plt.hist(s, bins=100) >>> plt.show() poisson(lam=1.0, size=None) Draw samples from a Poisson distribution. The Poisson distribution is the limit of the binomial distribution for large N. Parameters ---------- lam : float or array_like of floats Expectation of interval, must be >= 0. A sequence of expectation intervals must be broadcastable over the requested size. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``lam`` is a scalar. Otherwise, ``np.array(lam).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Poisson distribution. Notes ----- The Poisson distribution .. math:: f(k; \lambda)=\frac{\lambda^k e^{-\lambda}}{k!} For events with an expected separation :math:`\lambda` the Poisson distribution :math:`f(k; \lambda)` describes the probability of :math:`k` events occurring within the observed interval :math:`\lambda`. Because the output is limited to the range of the C int64 type, a ValueError is raised when `lam` is within 10 sigma of the maximum representable value. References ---------- .. [1] Weisstein, Eric W. "Poisson Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/PoissonDistribution.html .. [2] Wikipedia, "Poisson distribution", https://en.wikipedia.org/wiki/Poisson_distribution Examples -------- Draw samples from the distribution: >>> import numpy as np >>> rng = np.random.default_rng() >>> s = rng.poisson(5, 10000) Display histogram of the sample: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 14, density=True) >>> plt.show() Draw each 100 values for lambda 100 and 500: >>> s = rng.poisson(lam=(100., 500.), size=(100, 2)) laplace(loc=0.0, scale=1.0, size=None) Draw samples from the Laplace or double exponential distribution with specified location (or mean) and scale (decay). The Laplace distribution is similar to the Gaussian/normal distribution, but is sharper at the peak and has fatter tails. It represents the difference between two independent, identically distributed exponential random variables. Parameters ---------- loc : float or array_like of floats, optional The position, :math:`\mu`, of the distribution peak. Default is 0. scale : float or array_like of floats, optional :math:`\lambda`, the exponential decay. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Laplace distribution. Notes ----- It has the probability density function .. math:: f(x; \mu, \lambda) = \frac{1}{2\lambda} \exp\left(-\frac{|x - \mu|}{\lambda}\right). The first law of Laplace, from 1774, states that the frequency of an error can be expressed as an exponential function of the absolute magnitude of the error, which leads to the Laplace distribution. For many problems in economics and health sciences, this distribution seems to model the data better than the standard Gaussian distribution. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] Kotz, Samuel, et. al. "The Laplace Distribution and Generalizations, " Birkhauser, 2001. .. [3] Weisstein, Eric W. "Laplace Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/LaplaceDistribution.html .. [4] Wikipedia, "Laplace distribution", https://en.wikipedia.org/wiki/Laplace_distribution Examples -------- Draw samples from the distribution >>> loc, scale = 0., 1. >>> s = np.random.default_rng().laplace(loc, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 30, density=True) >>> x = np.arange(-8., 8., .01) >>> pdf = np.exp(-abs(x-loc)/scale)/(2.*scale) >>> plt.plot(x, pdf) Plot Gaussian for comparison: >>> g = (1/(scale * np.sqrt(2 * np.pi)) * ... np.exp(-(x - loc)**2 / (2 * scale**2))) >>> plt.plot(x,g) dirichlet(alpha, size=None) Draw samples from the Dirichlet distribution. Draw `size` samples of dimension k from a Dirichlet distribution. A Dirichlet-distributed random variable can be seen as a multivariate generalization of a Beta distribution. The Dirichlet distribution is a conjugate prior of a multinomial distribution in Bayesian inference. Parameters ---------- alpha : array Parameter of the distribution (k dimension for sample of dimension k). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- samples : ndarray, The drawn samples, of shape (size, alpha.ndim). Raises ------- ValueError If any value in alpha is less than or equal to zero Notes ----- The Dirichlet distribution is a distribution over vectors :math:`x` that fulfil the conditions :math:`x_i>0` and :math:`\sum_{i=1}^k x_i = 1`. The probability density function :math:`p` of a Dirichlet-distributed random vector :math:`X` is proportional to .. math:: p(x) \propto \prod_{i=1}^{k}{x^{\alpha_i-1}_i}, where :math:`\alpha` is a vector containing the positive concentration parameters. The method uses the following property for computation: let :math:`Y` be a random vector which has components that follow a standard gamma distribution, then :math:`X = \frac{1}{\sum_{i=1}^k{Y_i}} Y` is Dirichlet-distributed References ---------- .. [1] David McKay, "Information Theory, Inference and Learning Algorithms," chapter 23, http://www.inference.org.uk/mackay/itila/ .. [2] Wikipedia, "Dirichlet distribution", https://en.wikipedia.org/wiki/Dirichlet_distribution Examples -------- Taking an example cited in Wikipedia, this distribution can be used if one wanted to cut strings (each of initial length 1.0) into K pieces with different lengths, where each piece had, on average, a designated average length, but allowing some variation in the relative sizes of the pieces. >>> s = np.random.default_rng().dirichlet((10, 5, 3), 20).transpose() >>> import matplotlib.pyplot as plt >>> plt.barh(range(20), s[0]) >>> plt.barh(range(20), s[1], left=s[0], color='g') >>> plt.barh(range(20), s[2], left=s[0]+s[1], color='r') >>> plt.title("Lengths of Strings") Unsupported dtype "%s" for randomGenerator.triangular (line 2709)Generator.standard_t (line 1701)unable to allocate array data.probabilities do not sum to 1Generator.logseries (line 3375)Generator.lognormal (line 2460)Generator.geometric (line 3181)Generator.dirichlet (line 4027)Generator.chisquare (line 1488)Generator.vonmises (line 1795)Generator.rayleigh (line 2572)Generator.logistic (line 2380)Generator.binomial (line 2809)nsample must be nonnegative.itemsize <= 0 for cython.array'a' and 'p' must have same sizeGenerator.weibull (line 1976)Generator.shuffle (line 4166)Generator.poisson (line 3029)Generator.laplace (line 2176)Generator.integers (line 472)nsample must be an integermean must be 1 dimensionalRange exceeds valid boundsGenerator.uniform (line 871)Generator.pareto (line 1878)Generator.normal (line 1050)Generator.gumbel (line 2261)probabilities contain NaNnsample must not exceed %dInvalid shape in axis %d: %d.Generator.random (line 263)Generator.power (line 2075)Generator.gamma (line 1244)Generator.choice (line 630)Generator.zipf (line 3101)Generator.wald (line 2641)Generator.bytes (line 601)Cannot index with type '%s''p' must be 1-dimensionalnumpy.random._generatornumpy.core.multiarrayGenerator.f (line 1322)normalize_axis_indexNotImplementedErrornsample > sum(colors)ngood + nbad < nsamplecline_in_traceback__pyx_unpickle_Enumascontiguousarraymay_share_memory_bounded_integerssum(pvals[:-1]) > 1.0standard_normal__setstate_cython____pyx_PickleError_poisson_lam_maxdtype_is_objectallocate_bufferView.MemoryView_integers_types__generator_ctorRuntimeWarning__reduce_cython____pyx_getbuffer_generator.pyxcount_nonzerobit_generatorOverflowErrorstringsourcesearchsortedreturn_index__pyx_checksumRuntimeErrordefault_rngcheck_validPickleErrorMemoryErrorImportError__pyx_vtable____pyx_resultnumpy.dualmode > rightlogical_orless_equalleft == rightissubdtypeempty_likeValueErrorIndexError__reduce_ex____pyx_statemarginalsleft > modeenumerateTypeErrorGeneratorwarningsswapaxessubtract__setstate__reversed__reduce____pyx_typepickleoperatoritemsizeisscalarisnativeisfiniteintegers__getstate__floatingendpointcholeskyallcloseEllipsistobytesstridesshufflereshapereplacensamplememviewintegergreaterfortranfloat64float32capsule at 0x{:X}asarrayalpha <= 0updateunpackuniqueuint64uint32uint16structreducerandom_picklenamemethod__import__ignoreformatencodedoublecumsumcompatcolorsastypearangezerosuint8statestartsigmashapescalerightravelrangeraisepvals_pcg64numpyngoodkappaisnanint64int32int16indexflagsfinfoerrorequal__enter__emptydtypedfnumdfdencount__class__arrayalphaPCG64ASCIIwarn__test__takestopstepsqrtsortsizesideseedrtolprodpacknoncndimnbad__name__modemean__main__longlocklessleftitemintpint8highfull__exit__eigh__dict__copyboolbaseaxisatolzigtolsvd__str__outobj__new__lowloclamepsdotcovanyalladd333333?C?9B.? * ?,|l @yD@:5/?@@R2B@96SC@wz*E@r4dF@OOfq]@Ob^@+NT_@ݭC#`@~{`@kbba@YSȐa@n b@1Ib@5ca c@c@ͦ3 d@\>d@nz e@s9Je@FGGʪ f@yyuf@IJC g@Y&g@oFh@·h@aQL i@ai@ F~x*j@&Pj@7k@!+k@VFl@ l@tVm@pZNm@k9ihn@HQOUn@a,~|o@b4nʼnp@+e Ip@cp@)Vp@*q@6Gaq@q@>m#FJq@FK.5r@b)C|r@Wrr@V] s@rRs@GIqs@ >6qs@jB*t@ A=rt@fIw|t@d'-u@X+{ Mu@# u@ZGDu@;#(v@b%rv@iv{Իv@w@Ow@\&әw@}6-#w@h͙.x@k?7yx@–'x@_*y@Yy@1*y@^TTy@,{L ?A?Į?"?ʝ?G??i>l>7>>*J>>>^>>F>>7P>>K{>>>u>;->>>|>eO>4(>8>L>N>ȿ>>>>·>ε>߳>>>0>U>~>|>ަ>Y>IP>w>ҟ>B>e>)>~>_>>C>{>>J>﷍>X(>'>N>Í>x >b>x>>!>^}>;z>Хw>@t>wr>byl>i>g>Sd>3a>^>]\>&Y>z)W>T>P R>O>L>5~J>3H>E>nC>@>VK>>;>9>07>4>2>>0>p->+>s)>7'>%>"> >s>L>*>= >T>T>4>>y >ϣ > >>L>>l>=+==0 =C==8==hp==0== =n==|===;=ں=Z=o,=ް=ߗ=.W===%==r=W= C=4=M,=4*=D.=y8=H=~=x=3E3lM3FT3/[3b3i34p3fw3&~3[3B3ψ3g37!3>3T3d3n3r3Fq3j3_31P3r<3$3k 33ȸ3q3|{3P3#3C333dY3"3+3®3r35333x37333p3-33731b33l44(44h4C44 4` 4M 47 44?4nB44L4 i4a4T044542`44p. 4!4"4i$4%4@'4t(4>*4+4,4j.4/4'P1424):4454&)7484c:4;4$=4+>4@4A4KC4vD4B(F4G4:I4J4rTL4M4GuO4Q4R44T4U4EiW4Y4 Z4G\4]4_4:a4b4d4Bf4\g4ji4bk4m4n4p4\r4}"t4Yu4Hw4[y4X{46.}4 4q4a4]S4F4N<434,4+(4{%4$4o&4,*4'04m84 C4P4_4q474{4w4>ԕ44s4<4d444$4 (4a44lߢ4$4l44x 4_444{4 4EP4±4{:4귴4);4nķ4S444<*4տ44A4.44ע44f4RW4R4*Y4Fk44δ444444g44k4<444y44u4_45555@5ó 5 5]5^555q5v 5!5%5V*5s/5;S55:<5D5NO5^5Nv5QHqoMֻanjDotTrotou$w'xx,jyy7\z׻z{W{S{{.|3|]|ȃ|||I||}C0}F}Z}m}}S}(}}-}}"}}|}M}~i ~~~B(~o0~C8~?~F~M~T~Z~a~f~l~r~]w~v|~`~ ~~$~m~~~w~:~ަ~f~ѭ~#~Z~y~~q~K~~~^~~a~~~`~~~~~~~~f~*~~~-~~J~~=~~~\~~~$~U~}~~~~~pH`  i   6  H  A!B+m 5XttW3 `wK\ L   s   G {V~~~d~~x~~K~~~~~~~~~)~~~a~~~{~;~~A~~~m~~z~~~"~k~]~~~ԃ~|~s~j~Ua~W~K~?~2~$~~~ }}} }}i}A}}|Q|D|{3N{zeyww7ms?7E?P?'{{?*!?bv?mU?9U1T?/v?x]?&1$-?~ n?cK[!?I?\Omg?f?uLi=?sڂl?x?Qf?ij?%ᨯC?+?Dܻ?z?cE#;?^E#?$O?2m?P"K?>?{s?%;?omo?3;?J9?++?*T[?};1s?HeC?$`?vE!=?ſ-r?MBц?K=?Q}6Ei?7u? !?z}k? ~?@?`x?*?8? Qi?oTC?_(4?ָ?@je?!u v?7Zi?{ ?I?]T?9]??}?8aD:?Yζi?Ɲҷ?r^sSw?ꍰ07?d>[?%۹? Ə{?'HB>?vX#?l1&?:l?磽!O?ލ?&?ڋ?ﾀ+ j?A1?N0Z?0H?}G?(V?5$1!?pB9 ?b"FS?)vEW(?vG}rO?~ /? {^?Z? ?ބS?i"?lR?3Sn?>N?Ґ]b?,|y2?jG>?TLҫ?~>\O??@YH?/֎@?9O"H?>?1 7?8?Ox?]4?5D9g?r|?>ܸ$8? [B/?I䠟?O?y%d?bPޱ?c?PR?j?F}?9(Q1?c?(ڦ^w?0U^Q?1j?T ξ?x.BTv?Imb.??6YJ?)ِ?\C}?%d?w?SN?эv?pa?,Q&?@oű?SuFe?PV?;?I?viׯ?4D?.g?X1Iα?Jy?!dJ?پz?j»?8G;?L|{ʎ?mwn?k9:9??Ry?A&E?U?Ŗ<?k&_?G??~#? V#?_?S?Q| z? Y&?$?htQz? 3Tݜ?pXP?N梚?H*g?gS(u?1c?w@rT?Q=I?QA?]1%? RD?lj?W'n?-BU؊?h?t4? n?boQ?qvi?_)N?]tQW}?6H#z? 67w?"Ηs?C@Wi=q?ḰXl?f?$ka?%> T+Y? O?K 2=?]d<A]X`<+M[Ij<[5q&<.8eG< h#ឪV <;LC%K<ꆭh NVeΙVn6nvK zicp%E tQ)2U1WQ9Lin?23F:L"3\LQ V f[_rWDdx h+*k2=Ko:qr Mu\x?A{FS~8;b=ZV`bBtu9=JE>XدGwdO 8cx AFẙi&zqVYםΡag6 X83:뇡koɣj_ۤ| Mg^ݧt|Ψ_ΓXp2X^ttH蟿W;ޭl~$\z[߁İPp:J+N!X ɦ֬ ᆴX7(. Ɉ?5}h.G{tr&oya=cA/˺DH0⤮<)9O@ᣩTrVj֋@?˷dsI^i@(0߾ta&⊂l1EA1T[n&mi#d)B}QJwt}B < EOvpc/F<Ң"Ae އ0~ Rfq(*QtH3D@M`P}hwx%ƿ8*JG+[EliPIw+ E>ҙ02yΩ4A (Nt.Ȱ--̕^&܌z#;ޖu~g6X .pmF 3n bH޵LaEZvpR(-x_b˿ӰdyQӶVg<7܆ut7$MH𯋉ld"rqտH)݄ /0 wپ}2}K D5z&R cM,}uc?Ѡp5.bJ3ʸT[vv+\[U@ضBi"7oLeiFγ>SR(D2Z> B0$y1gWr-ެ @樫(afoeW-|&aY +M?V#z?u?q?}n?k?Lh?e?Rc?`?Zw^?*+\?Y?RW?U?_S?XQ?߱O?M?3K?J?GH?F?jD?`C?(`A?j??>?x,>N>>q>>>j>>k>>Π>>F>>>'>\>#>u>J>*>_F>d>+>$>w>>>JK>y>|>iݿ>>I>;>ʾ>t>5<> ~>>>O>>>~3>T>ե>(>g~>ՠ>G/>>>F>J> >:n>bԓ>Q<>>x>~>>>^>Ј>D>l>1>>%>\D>@|>?y>Bv>Hs>Qp>#^m>mj>|g>md>a>^>$[> Y>=3V>[S>P>M>J>~H>UE>B>?>=>S:>7>"4>=22>T/>d,>m+*>m'>c$>N?">,>>m>t>F>>1*> > >Y>>ʗ>>I=_={==^==&=_=g=='0===P6=˙=\= s==d= =yo=/=6=.=fЍ=x=i'=܀=a1y=p=xIh=_==W=TO=G=>=N6=.=&===-H==<א<̀<<<.4V?4=3@4A4A4qB4C4D4udE4-CF4K"G4H4H41I4J4vK4\fL4HM4+N4aO4O4bP4ٽQ4R4ԊS4crT4ZU4CV4-W4ZX4Y4UY4Z4[4(\4_]4^4_4C`4va4alb40cc47[d4~Te4Of4Jg42Hh4Fi4Fj4Hk4Kl4MPm4Vn4^o48hp4sq4r4s4 t4u4v4Cw4x4 z42{40S|4u}4~44v4@ 4L4>4ق4v444lV44R4F44p4 I44"4_44Ќ4l4L4`4ԏ4坐4y4ݖ4%44r&4k44(4444.4Q4N4t44\۶4H94̻4p44~X4w4p_4~444wE`mru\zw8xky5zz/ {ԃ{{7|3}|&|H|}C}g}ۇ}}a}g}]}~~4%~5~C~Q~g^~ij~u~>~2~~r~դ~Ƭ~N~u~C~~~~k~~~~~~t~~~6 < :#%](*.-z/13579;=?EABD:FGNIJ8LMNLPQR T=UdVWXYZ[\]^~__`;abbcod.eefLggh~~7~~/~7~~ ~ ~w~G]~>~Y~,}6}b}|O|06{x?yjD?l[T?w'??o?Wp?xI?-3?x^j??N?R:e?4:>?l?*?%z?PՋt?4?e;?$"?zaWF}?Gz‘B?Oq1? OU?ߺH?7a?nV,? K?Xhw?հ<?Vp\?m?)?zP?ZcX?*;Q^?#*'g? U7?e&$ ?jJo?\Ȭ)?L&?FS?leZ&?g ?NIO??xRr!?P_hy?y6IJO?_5%?[X~?1>?bU?+À?PX?5:pɗ0?8d?;U?J?͓?)m?ېZ]G?/|!? ?iT??Wq?PF9 ?ߓ^??ۮY?3???i?Z8o? O5?ٸ?P?R9?igP?La;?L?!ވ?%o?{7=8?Ҁt?DvC?6?=p\?;So&?mj?W?j?$O?z5Ѽ?Ҏ?C|P?yh|?%H?/ZM?f!w;??>ǭ?MAz?G?y?.?P9կ?TT}?g4K?#$O? Y?BM?6C;?B"_U?~t$?œ߉?52?Ҙl'?DɤT?<(i?qE8 ? Uī?OQM?o^?Sq͒?Gط5?zx?1zd}?:R!?Wg?~& ~k?=~-2?ZҿҶ?'|j_]?it?[?8R?uqb?#h?z|J?G~`?\!>?GF?vJ?l󈬚?5hȩmE?㭍?-l ?uG?1i%?調?M?e*|?zè?^V?4<%F?B}u?c-@c?n? R=?Kr?*}T#?,"k>?R) ?K{o?vaӽ?命8? t;I_? h?3xk?3Ӻ?b3?vZ9S?LJisk?M$a.?ftW?+ ?"@|?&#?p>_?1fҲ? DE?} ?/?%,?0?5nl+,&?QG?b. ?,*(>?p_8?cU)?h*?'wާ?dИۦ?ԭ<ڥ?]']ۤ?ݣ?=|?j?.?ĥׁ?u? ̓0?"NR? y? ڥ?d֔?^8 ?0`4I?IrO*?O'?x A?B?/)?7h`|?] ٨v?p?gC_e?T?yx;I< <[,L< Ŀk<4xV<=A[<'?}y<NG<~;[xo6xu{fUY>9>{ppCBwS(:5^dܓAN}8) YfHqն&|s f2,2Ztզޗ .n ZR'ӯB)[l@u Pҍ'TȈt(5wI'L/$;nXMØT`OArW,+jtȳRfARnqӊ<KZW$eKs) 4<=>)G'QA@Y.(5bX jz>lq{2Xx{~JH҄Cc`Qz%~ )Q\HsrUb'Bkq-hnק Ψ;3Kd)P^٨Tv$Hx"$ 5..&$ŗ: Aޓ=?~)@ lѿ3 ; @UUUUUU?"@m{?@@5gG8?SˆB?AAz?<ٰj_?$+K?88C?J?llf?dg?$@= ףp=@n?[ m?h|?5?333333 @r?$~?B>٬ @r鷯?Q?Q?9v?(\@ffffff@.@4@x&?@?UUUUUU?a@X@`@|@@MA>@@-DT! @h㈵>-DT!@3?r?q?0@;OZ )[ p( p Oq qltP\uuvwwXxxxHy`9yxTyyy z8zz({o{P{P}pDpӃV*8= 8 8p X 8 h (0 ( H x `  xH 8p  (H h  P8P8X(8Hh8X`x0XX88 PXhXH xX H(Hh((8hHX (!X&8(-@8.`/0158 9 H< x?X!B!xF"8Jp"xM"8Q@#T#X $x\$_$b`%f%i0&l&o&s`'v'z0(~(h)(p)h)(8*8h*H*8+H`+h++8 ,hp,,,0-Hp-ز-h-0.ȹx../XX/p///X00(1`1x 182"h2-2.27@383E3xF3JH4b4d4j855ض68X6(6H6xH7p778!8$@8$X8(8(88/8HH9o9x9x8:(:(:t(;x;H;0<8<h$<4P=X5h=5=5=5=H6>h7X>7>8>89(?9`?:?;?<(@=x@>@@@(C(AHC@AhCXACpACAXEAEAEAEBE BF8BGhB8GBGBGBGBHHBHC(I@CI`CICJChJCKCMDM8DXXDZDx[Ex\@E\`E]E`EaE8bFb Fb8FcXFdFHeFfGHgPGhGjPHjHnI(qIu JvJXwJxK{Kh~KzRx $F` FJ w?;*3$"D\4tEBDD d GBI AAB4ȷdAA ABY p(IDl A 4aDD A TP1TlBEE A(A0G@j 0A(A BBBI { 0A(A BBBE .$iPAK`A|[BBB B(D0D8D 8A0A(B BBBG D 8A0A(B BBBB b8K0A(B BBBDpBFB B(A0A8D`8A0A(B BBBDsBDA G0e  GABK H  AABA 4, [BFD D(H0|(A ABBdcA$(vD h AA DBEB D(D0G@z 0A(A BBBF ,,dcBID D0K AAB$$AG@ AF 4LdAGD0q AAA V AAH ؿjAw P a4(dAGD0q AAA V AAH `AT K w$ocGADA |ACL<cBBB B(D0A8DP8C0A(B BBBDfBEH D(D0G0A(A BBB$fhAKD0UAADfBLE E(D0A8DP8A0A(B BBBDDgBLE E(D0A8DP8A0A(B BBB4peBDD E GBH AAB>h$K,2hfBDD XAB$$hh?AGA pDA4LhZBGD ~ ABE CABh$h$h$hWAGD0EDAhAZD -AD A AI G AH [ AL G AH ^ AI ThDAG tCth4`wAAG0[ CAH v CAF hAD A,iYBDG D0C AAB HiGAD @A4 8z<L AD h AJ Q AF D CI G AA 4 iBEG D(D@p(A ABB, liTBAA LABL j BAA G0  DABB D  DABE E DABLD `lBEB B(A0E8G8A0A(B BBB (ZG R nAG P X  P  H , @ \D nBBA D(E0{ (A ABBE ^ (G ABBE A(A ABBL ؾBBA D(G0X (A ABBE V (A ABBA 4 oBBA A(A0u(A ABB, ^oGI VE b$L oWAGD0EDA$t pAG@ AI  CtNL ( BEB B(A0A8G 8A0A(B BBBA , TADD ^ DAA 44 uBDD Z GBC AAB$l `cADD UCAD NBAC G0R  AABE T  AABH 4 nADD K AAJ DDA4fADD I AAD DDA$L ADD CALtBBB E(D0A8GP 8A0A(B BBBG LBEB E(D0C8D@z 8D0A(B BBBH ,3BDB B(A0A8G` 8A0A(B BBBE  8A0A(B BBDB  8F0A(B BBBI \ 8F0A(B BBBA 8BBB E(D0H8DP 8A0A(B BBBH d 8A0A(B BBBJ u 8D0A(B BBBF B 8E0A(B BBBH ,\AAAG  CAF LBEB B(A0A8G 8A0A(B BBBA $p5AG0 AE LBEE B(A0A8Gp 8D0A(B BBBE <TOBBA A(D@ (A ABBI <BED D(G` (A ABBF TXBAA z ABG D ABI A OBN A DBE ,RBBB E(D0C8F 8A0A(B BBBF , 8D0A(B BBBG  8C0A(B BBBE ,BDA  ABK LBGB B(A0C8DPS 8D0A(B BBBK ,4HZAAG0 AAA dxkD R A kD R A <aBBA A(G@ (A ABBA <H)BBA A(G0 (D ABBA <$8BBA A(I0 (D ABBA <daBDA  DBA } DBA ,AJ AB R AA ,pAJ AB R AA ,4AAD  DAD ,dADG J AAA LhBEH B(A0A8J 8A0A(B BBBE DBBB D(D0GPF 0A(A BBBA L,@ BBB J(A0A8D@ 8D0A(B BBBI |hD O A ,HBDD Q ABB ,BDD P ABC , ADD0q AAA D,6BBB D(D0GP# 0A(A BBBH \t(BBB D(D0G 0A(A BBBA  0A(A BBBG \x(BBB D(D0G 0A(A BBBA  0A(A BBBG \4H(BBB D(D0G 0A(A BBBA  0A(A BBBG L3BBA D(G (A ABBF  (A ABBF \!BBB D(D0G 0A(A BBBA  0A(A BBBI dDh$BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBG \'3BBB D(A0G 0A(A BBBB  0A(A BBBE l *BBB D(A0G 0A(A BBBD  0C(A BBBG V 0A(A BBBB l|-BBB D(A0G 0A(A BBBD  0C(A BBBG V 0A(A BBBB l@1BBB D(A0G 0A(A BBBD  0C(A BBBG V 0A(A BBBB l\4BBB D(A0G 0A(A BBBD  0C(A BBBG V 0A(A BBBB d7(BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBE d4:(BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBE dp=(BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBE d8@BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBG dlC(BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBE \XFBBB D(A0G 0A(A BBBA  0A(A BBBH d4HBBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBG d L(BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBE dN=BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBG dlRBBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBG dVBBB E(D0A8G 8A0A(B BBBI  8A0A(B BBBH l< `YBBB D(A0G 0A(A BBBK  0C(A BBBG V 0A(A BBBB \ \3BBB D(A0G 0A(A BBBB  0A(A BBBE d !_BBB E(D0A8G 8A0A(B BBBA  8A0A(B BBBG ,t!bADD i AAD $!c AD s AA ,!dBHD  ABD L!,Z:BBB B(A0A8J:8A0A(B BBBLL" hBEB E(D0A8Dp 8A0A(B BBBG ,"nADD0x AAA L"o!BEB E(D0A8Dp 8A0A(B BBBE <#vBBJ A(A0= (A BBBD L\#|!BED A(G@ (D ABBG l (D ABBK <#}BBA A(D0 (C ABBE <#@BBA A(D0 (C ABBE <,$BBA A(D0 (C ABBE <l$/BBA I(D0 (D ABBD <$ЃBBA A(D0 (C ABBE <$ BBA A(D0 (C ABBE <,%pBBA A(D0 (C ABBE Dl%BBB D(A0D@ 0D(A BBBF <%H>BBE A(G0 (D BBBA L%HBBB B(A0A8GP 8A0A(B BBBE LD&BEB E(A0A8D` 8D0A(B BBBA &RoYL&@OBBB B(A0D8D` 8A0A(B BBBG ,&@ADJ AAA d,' BEB E(A0A8GA 8C0A(B BBBD  8D0A(B BBBA d' BEB E(A0A8GA 8C0A(B BBBD  8D0A(B BBBA L'` BBB B(A0A8Dp 8D0A(B BBBD LL(BBB E(A0A8DP 8A0A(B BBBJ L(P BBB E(D0D8Gc 8A0A(B BBBI d(BEB E(D0A8D 8D0A(B BBBA  8C0A(B BBBD LT)BBB B(D0D8D 8A0A(B BBBE L)x BBG E(A0A8G 8D0A(B BBBG )8DZ B d*BBB B(A0A8Gp 8D0A(B BBBA  8C0A(B BBBA D|*PpDA D0W  AABD hW0L*( BBG B(A0D8G 8D0A(B BBBG +Dj B L4+x+BBB E(A0D8GP 8F0A(B BBBH L+X`BLB B(A0A8G  8D0A(B BBBA L+h-BIB A(A0 (A BBBE  (A BBBA L$,0BBB B(A0D8DPu 8A0A(B BBBH dt,X5[&BEB B(A0D8Gd 8C0A(B BBBI  8D0A(B BBBA d,P[%BBB B(D0A8G 8C0A(B BBBF  8A0A(B BBBJ LD-ȀZBBB E(A0A8D 8D0A(B BBBJ L-؈(BBE B(A0D8J 8A0A(B BBBA L-xBBB B(D0A8G@ 8F0A(B BBBE L4.H#BBB E(E0E8F@ 8A0A(B BBBK $.(GAG AK L.P)BDB B(A0C8I ` 8A0A(B BBBG L.BDB B(I0A8JQ 8A0A(B BBBA ,L/ AAJ AAI |/4/BAA J  AABD /L/BEE B(A0D8G 8A0A(B BBBF L408\BBB B(A0D8Ga 8A0A(B BBBA L0H 'BBB B(D0A8JK 8A0A(B BBBD L06ABBB B(D0D8G 8A0A(B BBBB L$1w*BBB B(A0D8J 8A0A(B BBBD Lt18 BBB B(A0A8G`K 8A0A(B BBBJ L1BEB B(D0A8G* 8A0A(B BBBA L2HlpBEB B(A0D8J 8A0A(B BBBC Ld2h9BDB E(A0A8G5F 8D0A(B BBBH d28GBBE E(A0A8G 8A0A(B BBBD  8C0A(B BBBD L3`K BBB B(A0D8Js 8A0A(B BBBD Ll3k',BBB B(A0D8J 8A0A(B BBBA L3/PBBB B(A0D8J 8A0A(B BBBF d 4PBBB B(A0D8G 8D0A(B BBBH  8C0A(B BBBB t4W 4`a m4"GZ4 44BBED D(F0f(A ABB4 5RBED D(F0v(A ABBLD58BHH G@m  EABB   AAFD k AAF45BBED D(F0f(A ABB\5 )BBH H(G@g (D ABBE  (A ABED v(A ABF4,6BBED D(F0f(A ABB4d6_BBD A(D0K(A ABB460|BBD A(D0h(A ABBT6xBBI H(K0GP 0A(A BBBD 0A(A BBB4,7BBED D(F0f(A ABBLd7BBH H(G@ (A ABBE (A ABB47HBBED D(F0f(A ABB<7`nAG@\ AK  AA N AA M AE 4,8AG0l AA   AA J AE d8DM|8DM8DM8 48vAD@M EAB `8(D c9D U,9,D gD9D U\9D U,t9AG0D AS A9DQ9WAK }E9/AG ]A9'D b:FH P H ]4:HCD r,L:AG R AM XA|:kAG YA:0NAG DA:`DI:hTD @ E $:AG W AI 4; AAGpe AAH T DAG T; &IXLt; [ BBB E(D0A8J 8A0A(B BBEA <;eBID D(Dpr (A ABBA 4<D@ L s E C E A O L<<<pAG  AD L AC K DI h AE |<0[AK AE$<pEO0 EA <<8AGP AG L AK ^ AA  AE $=AG@ AG ,=WD RD=(`D V\=pt=xAK0F$=XH@ I Z A ,= BFD c ABF L=NBBE B(D0D8DPw 8D0A(B BBBE L<>BBE E(A0C8D@o 8D0A(B BBBF |>BDE B(A0A8DPq 8A0A(B BBBB \ 8A0A(B BBBJ  8D0A(B BFBK | ?!BDE B(A0A8DPm 8A0A(B BBBF [ 8A0A(B BBBK  8E0A(B BFBJ ,?0"^AF M AK S AL |?`"BBB E(A0D8G` 8A0A(B BBBF  8A0A(B BBBB  8A0A(B BBBA <@%BBB E(D0A8FP 8A0A(B BBBD { 8A0A(B BBBA e 8A0A(B BBBA o 8A0A(B BBBA @'EB B(A0C8DP8A0A(B BBBDP 8A0A(B BBBE ^\\A*`PV8A0A(B BBBJBB E(A0C8DP4A`,fbID _ABG \A,BEB B(E0D8D`p 8A0A(B BBBA [8A0A(B BBBdTB,BDB B(A0A8Dh 8A0A(B BBBA   8C0A(B BBBF LBP0BBB B(A0D8D 8A0A(B BBBB L C2BBB E(A0E8G 8D0A(B BBBH 0.-)ak{ 8 ))opH  )@4w oH4oo2o )fvƽֽ&6FVfvƾ־&6FVfvƿֿ&6FVfv&6FVfv&6FVfv&6FVfv&6FVfv&6FVfv&6FVfv&6FVfv&6FVfvConstruct a new Generator with the default BitGenerator (PCG64). Parameters ---------- seed : {None, int, array_like[ints], SeedSequence, BitGenerator, Generator}, optional A seed to initialize the `BitGenerator`. If None, then fresh, unpredictable entropy will be pulled from the OS. If an ``int`` or ``array_like[ints]`` is passed, then it will be passed to `SeedSequence` to derive the initial `BitGenerator` state. One may also pass in a`SeedSequence` instance Additionally, when passed a `BitGenerator`, it will be wrapped by `Generator`. If passed a `Generator`, it will be returned unaltered. Returns ------- Generator The initialized generator object. Notes ----- If ``seed`` is not a `BitGenerator` or a `Generator`, a new `BitGenerator` is instantiated. This function does not manage a default global instance. permutation(x, axis=0) Randomly permute a sequence, or return a permuted range. Parameters ---------- x : int or array_like If `x` is an integer, randomly permute ``np.arange(x)``. If `x` is an array, make a copy and shuffle the elements randomly. axis : int, optional The axis which `x` is shuffled along. Default is 0. Returns ------- out : ndarray Permuted sequence or array range. Examples -------- >>> rng = np.random.default_rng() >>> rng.permutation(10) array([1, 7, 4, 3, 0, 9, 2, 5, 8, 6]) # random >>> rng.permutation([1, 4, 9, 12, 15]) array([15, 1, 9, 4, 12]) # random >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr) array([[6, 7, 8], # random [0, 1, 2], [3, 4, 5]]) >>> rng.permutation("abc") Traceback (most recent call last): ... numpy.AxisError: x must be an integer or at least 1-dimensional >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr, axis=1) array([[0, 2, 1], # random [3, 5, 4], [6, 8, 7]]) shuffle(x, axis=0) Modify a sequence in-place by shuffling its contents. The order of sub-arrays is changed but their contents remains the same. Parameters ---------- x : array_like The array or list to be shuffled. axis : int, optional The axis which `x` is shuffled along. Default is 0. It is only supported on `ndarray` objects. Returns ------- None Examples -------- >>> rng = np.random.default_rng() >>> arr = np.arange(10) >>> rng.shuffle(arr) >>> arr [1 7 5 2 9 4 3 6 0 8] # random >>> arr = np.arange(9).reshape((3, 3)) >>> rng.shuffle(arr) >>> arr array([[3, 4, 5], # random [6, 7, 8], [0, 1, 2]]) >>> arr = np.arange(9).reshape((3, 3)) >>> rng.shuffle(arr, axis=1) >>> arr array([[2, 0, 1], # random [5, 3, 4], [8, 6, 7]]) dirichlet(alpha, size=None) Draw samples from the Dirichlet distribution. Draw `size` samples of dimension k from a Dirichlet distribution. A Dirichlet-distributed random variable can be seen as a multivariate generalization of a Beta distribution. The Dirichlet distribution is a conjugate prior of a multinomial distribution in Bayesian inference. Parameters ---------- alpha : array Parameter of the distribution (k dimension for sample of dimension k). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- samples : ndarray, The drawn samples, of shape (size, alpha.ndim). Raises ------- ValueError If any value in alpha is less than or equal to zero Notes ----- The Dirichlet distribution is a distribution over vectors :math:`x` that fulfil the conditions :math:`x_i>0` and :math:`\sum_{i=1}^k x_i = 1`. The probability density function :math:`p` of a Dirichlet-distributed random vector :math:`X` is proportional to .. math:: p(x) \propto \prod_{i=1}^{k}{x^{\alpha_i-1}_i}, where :math:`\alpha` is a vector containing the positive concentration parameters. The method uses the following property for computation: let :math:`Y` be a random vector which has components that follow a standard gamma distribution, then :math:`X = \frac{1}{\sum_{i=1}^k{Y_i}} Y` is Dirichlet-distributed References ---------- .. [1] David McKay, "Information Theory, Inference and Learning Algorithms," chapter 23, http://www.inference.org.uk/mackay/itila/ .. [2] Wikipedia, "Dirichlet distribution", https://en.wikipedia.org/wiki/Dirichlet_distribution Examples -------- Taking an example cited in Wikipedia, this distribution can be used if one wanted to cut strings (each of initial length 1.0) into K pieces with different lengths, where each piece had, on average, a designated average length, but allowing some variation in the relative sizes of the pieces. >>> s = np.random.default_rng().dirichlet((10, 5, 3), 20).transpose() >>> import matplotlib.pyplot as plt >>> plt.barh(range(20), s[0]) >>> plt.barh(range(20), s[1], left=s[0], color='g') >>> plt.barh(range(20), s[2], left=s[0]+s[1], color='r') >>> plt.title("Lengths of Strings") multivariate_hypergeometric(colors, nsample, size=None, method='marginals') Generate variates from a multivariate hypergeometric distribution. The multivariate hypergeometric distribution is a generalization of the hypergeometric distribution. Choose ``nsample`` items at random without replacement from a collection with ``N`` distinct types. ``N`` is the length of ``colors``, and the values in ``colors`` are the number of occurrences of that type in the collection. The total number of items in the collection is ``sum(colors)``. Each random variate generated by this function is a vector of length ``N`` holding the counts of the different types that occurred in the ``nsample`` items. The name ``colors`` comes from a common description of the distribution: it is the probability distribution of the number of marbles of each color selected without replacement from an urn containing marbles of different colors; ``colors[i]`` is the number of marbles in the urn with color ``i``. Parameters ---------- colors : sequence of integers The number of each type of item in the collection from which a sample is drawn. The values in ``colors`` must be nonnegative. To avoid loss of precision in the algorithm, ``sum(colors)`` must be less than ``10**9`` when `method` is "marginals". nsample : int The number of items selected. ``nsample`` must not be greater than ``sum(colors)``. size : int or tuple of ints, optional The number of variates to generate, either an integer or a tuple holding the shape of the array of variates. If the given size is, e.g., ``(k, m)``, then ``k * m`` variates are drawn, where one variate is a vector of length ``len(colors)``, and the return value has shape ``(k, m, len(colors))``. If `size` is an integer, the output has shape ``(size, len(colors))``. Default is None, in which case a single variate is returned as an array with shape ``(len(colors),)``. method : string, optional Specify the algorithm that is used to generate the variates. Must be 'count' or 'marginals' (the default). See the Notes for a description of the methods. Returns ------- variates : ndarray Array of variates drawn from the multivariate hypergeometric distribution. See Also -------- hypergeometric : Draw samples from the (univariate) hypergeometric distribution. Notes ----- The two methods do not return the same sequence of variates. The "count" algorithm is roughly equivalent to the following numpy code:: choices = np.repeat(np.arange(len(colors)), colors) selection = np.random.choice(choices, nsample, replace=False) variate = np.bincount(selection, minlength=len(colors)) The "count" algorithm uses a temporary array of integers with length ``sum(colors)``. The "marginals" algorithm generates a variate by using repeated calls to the univariate hypergeometric sampler. It is roughly equivalent to:: variate = np.zeros(len(colors), dtype=np.int64) # `remaining` is the cumulative sum of `colors` from the last # element to the first; e.g. if `colors` is [3, 1, 5], then # `remaining` is [9, 6, 5]. remaining = np.cumsum(colors[::-1])[::-1] for i in range(len(colors)-1): if nsample < 1: break variate[i] = hypergeometric(colors[i], remaining[i+1], nsample) nsample -= variate[i] variate[-1] = nsample The default method is "marginals". For some cases (e.g. when `colors` contains relatively small integers), the "count" method can be significantly faster than the "marginals" method. If performance of the algorithm is important, test the two methods with typical inputs to decide which works best. .. versionadded:: 1.18.0 Examples -------- >>> colors = [16, 8, 4] >>> seed = 4861946401452 >>> gen = np.random.Generator(np.random.PCG64(seed)) >>> gen.multivariate_hypergeometric(colors, 6) array([5, 0, 1]) >>> gen.multivariate_hypergeometric(colors, 6, size=3) array([[5, 0, 1], [2, 2, 2], [3, 3, 0]]) >>> gen.multivariate_hypergeometric(colors, 6, size=(2, 2)) array([[[3, 2, 1], [3, 2, 1]], [[4, 1, 1], [3, 2, 1]]]) multinomial(n, pvals, size=None) Draw samples from a multinomial distribution. The multinomial distribution is a multivariate generalization of the binomial distribution. Take an experiment with one of ``p`` possible outcomes. An example of such an experiment is throwing a dice, where the outcome can be 1 through 6. Each sample drawn from the distribution represents `n` such experiments. Its values, ``X_i = [X_0, X_1, ..., X_p]``, represent the number of times the outcome was ``i``. Parameters ---------- n : int or array-like of ints Number of experiments. pvals : sequence of floats, length p Probabilities of each of the ``p`` different outcomes. These must sum to 1 (however, the last element is always assumed to account for the remaining probability, as long as ``sum(pvals[:-1]) <= 1)``. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- out : ndarray The drawn samples, of shape *size*, if that was provided. If not, the shape is ``(N,)``. In other words, each entry ``out[i,j,...,:]`` is an N-dimensional value drawn from the distribution. Examples -------- Throw a dice 20 times: >>> rng = np.random.default_rng() >>> rng.multinomial(20, [1/6.]*6, size=1) array([[4, 1, 7, 5, 2, 1]]) # random It landed 4 times on 1, once on 2, etc. Now, throw the dice 20 times, and 20 times again: >>> rng.multinomial(20, [1/6.]*6, size=2) array([[3, 4, 3, 3, 4, 3], [2, 4, 3, 4, 0, 7]]) # random For the first run, we threw 3 times 1, 4 times 2, etc. For the second, we threw 2 times 1, 4 times 2, etc. Now, do one experiment throwing the dice 10 time, and 10 times again, and another throwing the dice 20 times, and 20 times again: >>> rng.multinomial([[10], [20]], [1/6.]*6, size=2) array([[[2, 4, 0, 1, 2, 1], [1, 3, 0, 3, 1, 2]], [[1, 4, 4, 4, 4, 3], [3, 3, 2, 5, 5, 2]]]) # random The first array shows the outcomes of throwing the dice 10 times, and the second shows the outcomes from throwing the dice 20 times. A loaded die is more likely to land on number 6: >>> rng.multinomial(100, [1/7.]*5 + [2/7.]) array([11, 16, 14, 17, 16, 26]) # random The probability inputs should be normalized. As an implementation detail, the value of the last entry is ignored and assumed to take up any leftover probability mass, but this should not be relied on. A biased coin which has twice as much weight on one side as on the other should be sampled like so: >>> rng.multinomial(100, [1.0 / 3, 2.0 / 3]) # RIGHT array([38, 62]) # random not like: >>> rng.multinomial(100, [1.0, 2.0]) # WRONG Traceback (most recent call last): ValueError: pvals < 0, pvals > 1 or pvals contains NaNs multivariate_normal(mean, cov, size=None, check_valid='warn', tol=1e-8) Draw random samples from a multivariate normal distribution. The multivariate normal, multinormal or Gaussian distribution is a generalization of the one-dimensional normal distribution to higher dimensions. Such a distribution is specified by its mean and covariance matrix. These parameters are analogous to the mean (average or "center") and variance (standard deviation, or "width," squared) of the one-dimensional normal distribution. Parameters ---------- mean : 1-D array_like, of length N Mean of the N-dimensional distribution. cov : 2-D array_like, of shape (N, N) Covariance matrix of the distribution. It must be symmetric and positive-semidefinite for proper sampling. size : int or tuple of ints, optional Given a shape of, for example, ``(m,n,k)``, ``m*n*k`` samples are generated, and packed in an `m`-by-`n`-by-`k` arrangement. Because each sample is `N`-dimensional, the output shape is ``(m,n,k,N)``. If no shape is specified, a single (`N`-D) sample is returned. check_valid : { 'warn', 'raise', 'ignore' }, optional Behavior when the covariance matrix is not positive semidefinite. tol : float, optional Tolerance when checking the singular values in covariance matrix. cov is cast to double before the check. method : { 'svd', 'eigh', 'cholesky'}, optional The cov input is used to compute a factor matrix A such that ``A @ A.T = cov``. This argument is used to select the method used to compute the factor matrix A. The default method 'svd' is the slowest, while 'cholesky' is the fastest but less robust than the slowest method. The method `eigh` uses eigen decomposition to compute A and is faster than svd but slower than cholesky. .. versionadded:: 1.18.0 Returns ------- out : ndarray The drawn samples, of shape *size*, if that was provided. If not, the shape is ``(N,)``. In other words, each entry ``out[i,j,...,:]`` is an N-dimensional value drawn from the distribution. Notes ----- The mean is a coordinate in N-dimensional space, which represents the location where samples are most likely to be generated. This is analogous to the peak of the bell curve for the one-dimensional or univariate normal distribution. Covariance indicates the level to which two variables vary together. From the multivariate normal distribution, we draw N-dimensional samples, :math:`X = [x_1, x_2, ... x_N]`. The covariance matrix element :math:`C_{ij}` is the covariance of :math:`x_i` and :math:`x_j`. The element :math:`C_{ii}` is the variance of :math:`x_i` (i.e. its "spread"). Instead of specifying the full covariance matrix, popular approximations include: - Spherical covariance (`cov` is a multiple of the identity matrix) - Diagonal covariance (`cov` has non-negative elements, and only on the diagonal) This geometrical property can be seen in two dimensions by plotting generated data-points: >>> mean = [0, 0] >>> cov = [[1, 0], [0, 100]] # diagonal covariance Diagonal covariance means that points are oriented along x or y-axis: >>> import matplotlib.pyplot as plt >>> x, y = np.random.default_rng().multivariate_normal(mean, cov, 5000).T >>> plt.plot(x, y, 'x') >>> plt.axis('equal') >>> plt.show() Note that the covariance matrix must be positive semidefinite (a.k.a. nonnegative-definite). Otherwise, the behavior of this method is undefined and backwards compatibility is not guaranteed. References ---------- .. [1] Papoulis, A., "Probability, Random Variables, and Stochastic Processes," 3rd ed., New York: McGraw-Hill, 1991. .. [2] Duda, R. O., Hart, P. E., and Stork, D. G., "Pattern Classification," 2nd ed., New York: Wiley, 2001. Examples -------- >>> mean = (1, 2) >>> cov = [[1, 0], [0, 1]] >>> rng = np.random.default_rng() >>> x = rng.multivariate_normal(mean, cov, (3, 3)) >>> x.shape (3, 3, 2) We can use a different method other than the default to factorize cov: >>> y = rng.multivariate_normal(mean, cov, (3, 3), method='cholesky') >>> y.shape (3, 3, 2) The following is probably true, given that 0.6 is roughly twice the standard deviation: >>> list((x[0,0,:] - mean) < 0.6) [True, True] # random logseries(p, size=None) Draw samples from a logarithmic series distribution. Samples are drawn from a log series distribution with specified shape parameter, 0 < ``p`` < 1. Parameters ---------- p : float or array_like of floats Shape parameter for the distribution. Must be in the range (0, 1). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logarithmic series distribution. See Also -------- scipy.stats.logser : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function for the Log Series distribution is .. math:: P(k) = \frac{-p^k}{k \ln(1-p)}, where p = probability. The log series distribution is frequently used to represent species richness and occurrence, first proposed by Fisher, Corbet, and Williams in 1943 [2]. It may also be used to model the numbers of occupants seen in cars [3]. References ---------- .. [1] Buzas, Martin A.; Culver, Stephen J., Understanding regional species diversity through the log series distribution of occurrences: BIODIVERSITY RESEARCH Diversity & Distributions, Volume 5, Number 5, September 1999 , pp. 187-195(9). .. [2] Fisher, R.A,, A.S. Corbet, and C.B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology, 12:42-58. .. [3] D. J. Hand, F. Daly, D. Lunn, E. Ostrowski, A Handbook of Small Data Sets, CRC Press, 1994. .. [4] Wikipedia, "Logarithmic distribution", https://en.wikipedia.org/wiki/Logarithmic_distribution Examples -------- Draw samples from the distribution: >>> a = .6 >>> s = np.random.default_rng().logseries(a, 10000) >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s) # plot against distribution >>> def logseries(k, p): ... return -p**k/(k*np.log(1-p)) >>> plt.plot(bins, logseries(bins, a) * count.max()/ ... logseries(bins, a).max(), 'r') >>> plt.show() hypergeometric(ngood, nbad, nsample, size=None) Draw samples from a Hypergeometric distribution. Samples are drawn from a hypergeometric distribution with specified parameters, `ngood` (ways to make a good selection), `nbad` (ways to make a bad selection), and `nsample` (number of items sampled, which is less than or equal to the sum ``ngood + nbad``). Parameters ---------- ngood : int or array_like of ints Number of ways to make a good selection. Must be nonnegative and less than 10**9. nbad : int or array_like of ints Number of ways to make a bad selection. Must be nonnegative and less than 10**9. nsample : int or array_like of ints Number of items sampled. Must be nonnegative and less than ``ngood + nbad``. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if `ngood`, `nbad`, and `nsample` are all scalars. Otherwise, ``np.broadcast(ngood, nbad, nsample).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized hypergeometric distribution. Each sample is the number of good items within a randomly selected subset of size `nsample` taken from a set of `ngood` good items and `nbad` bad items. See Also -------- multivariate_hypergeometric : Draw samples from the multivariate hypergeometric distribution. scipy.stats.hypergeom : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Hypergeometric distribution is .. math:: P(x) = \frac{\binom{g}{x}\binom{b}{n-x}}{\binom{g+b}{n}}, where :math:`0 \le x \le n` and :math:`n-b \le x \le g` for P(x) the probability of ``x`` good results in the drawn sample, g = `ngood`, b = `nbad`, and n = `nsample`. Consider an urn with black and white marbles in it, `ngood` of them are black and `nbad` are white. If you draw `nsample` balls without replacement, then the hypergeometric distribution describes the distribution of black balls in the drawn sample. Note that this distribution is very similar to the binomial distribution, except that in this case, samples are drawn without replacement, whereas in the Binomial case samples are drawn with replacement (or the sample space is infinite). As the sample space becomes large, this distribution approaches the binomial. The arguments `ngood` and `nbad` each must be less than `10**9`. For extremely large arguments, the algorithm that is used to compute the samples [4]_ breaks down because of loss of precision in floating point calculations. For such large values, if `nsample` is not also large, the distribution can be approximated with the binomial distribution, `binomial(n=nsample, p=ngood/(ngood + nbad))`. References ---------- .. [1] Lentner, Marvin, "Elementary Applied Statistics", Bogden and Quigley, 1972. .. [2] Weisstein, Eric W. "Hypergeometric Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/HypergeometricDistribution.html .. [3] Wikipedia, "Hypergeometric distribution", https://en.wikipedia.org/wiki/Hypergeometric_distribution .. [4] Stadlober, Ernst, "The ratio of uniforms approach for generating discrete random variates", Journal of Computational and Applied Mathematics, 31, pp. 181-189 (1990). Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> ngood, nbad, nsamp = 100, 2, 10 # number of good, number of bad, and number of samples >>> s = rng.hypergeometric(ngood, nbad, nsamp, 1000) >>> from matplotlib.pyplot import hist >>> hist(s) # note that it is very unlikely to grab both bad items Suppose you have an urn with 15 white and 15 black marbles. If you pull 15 marbles at random, how likely is it that 12 or more of them are one color? >>> s = rng.hypergeometric(15, 15, 15, 100000) >>> sum(s>=12)/100000. + sum(s<=3)/100000. # answer = 0.003 ... pretty unlikely! geometric(p, size=None) Draw samples from the geometric distribution. Bernoulli trials are experiments with one of two outcomes: success or failure (an example of such an experiment is flipping a coin). The geometric distribution models the number of trials that must be run in order to achieve success. It is therefore supported on the positive integers, ``k = 1, 2, ...``. The probability mass function of the geometric distribution is .. math:: f(k) = (1 - p)^{k - 1} p where `p` is the probability of success of an individual trial. Parameters ---------- p : float or array_like of floats The probability of success of an individual trial. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized geometric distribution. Examples -------- Draw ten thousand values from the geometric distribution, with the probability of an individual success equal to 0.35: >>> z = np.random.default_rng().geometric(p=0.35, size=10000) How many trials succeeded after a single run? >>> (z == 1).sum() / 10000. 0.34889999999999999 #random zipf(a, size=None) Draw samples from a Zipf distribution. Samples are drawn from a Zipf distribution with specified parameter `a` > 1. The Zipf distribution (also known as the zeta distribution) is a continuous probability distribution that satisfies Zipf's law: the frequency of an item is inversely proportional to its rank in a frequency table. Parameters ---------- a : float or array_like of floats Distribution parameter. Must be greater than 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Zipf distribution. See Also -------- scipy.stats.zipf : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability density for the Zipf distribution is .. math:: p(x) = \frac{x^{-a}}{\zeta(a)}, where :math:`\zeta` is the Riemann Zeta function. It is named for the American linguist George Kingsley Zipf, who noted that the frequency of any word in a sample of a language is inversely proportional to its rank in the frequency table. References ---------- .. [1] Zipf, G. K., "Selected Studies of the Principle of Relative Frequency in Language," Cambridge, MA: Harvard Univ. Press, 1932. Examples -------- Draw samples from the distribution: >>> a = 2. # parameter >>> s = np.random.default_rng().zipf(a, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy import special # doctest: +SKIP Truncate s values at 50 so plot is interesting: >>> count, bins, ignored = plt.hist(s[s<50], ... 50, density=True) >>> x = np.arange(1., 50.) >>> y = x**(-a) / special.zetac(a) # doctest: +SKIP >>> plt.plot(x, y/max(y), linewidth=2, color='r') # doctest: +SKIP >>> plt.show() poisson(lam=1.0, size=None) Draw samples from a Poisson distribution. The Poisson distribution is the limit of the binomial distribution for large N. Parameters ---------- lam : float or array_like of floats Expectation of interval, must be >= 0. A sequence of expectation intervals must be broadcastable over the requested size. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``lam`` is a scalar. Otherwise, ``np.array(lam).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Poisson distribution. Notes ----- The Poisson distribution .. math:: f(k; \lambda)=\frac{\lambda^k e^{-\lambda}}{k!} For events with an expected separation :math:`\lambda` the Poisson distribution :math:`f(k; \lambda)` describes the probability of :math:`k` events occurring within the observed interval :math:`\lambda`. Because the output is limited to the range of the C int64 type, a ValueError is raised when `lam` is within 10 sigma of the maximum representable value. References ---------- .. [1] Weisstein, Eric W. "Poisson Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/PoissonDistribution.html .. [2] Wikipedia, "Poisson distribution", https://en.wikipedia.org/wiki/Poisson_distribution Examples -------- Draw samples from the distribution: >>> import numpy as np >>> rng = np.random.default_rng() >>> s = rng.poisson(5, 10000) Display histogram of the sample: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 14, density=True) >>> plt.show() Draw each 100 values for lambda 100 and 500: >>> s = rng.poisson(lam=(100., 500.), size=(100, 2)) negative_binomial(n, p, size=None) Draw samples from a negative binomial distribution. Samples are drawn from a negative binomial distribution with specified parameters, `n` successes and `p` probability of success where `n` is > 0 and `p` is in the interval [0, 1]. Parameters ---------- n : float or array_like of floats Parameter of the distribution, > 0. p : float or array_like of floats Parameter of the distribution, >= 0 and <=1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``n`` and ``p`` are both scalars. Otherwise, ``np.broadcast(n, p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized negative binomial distribution, where each sample is equal to N, the number of failures that occurred before a total of n successes was reached. Notes ----- The probability mass function of the negative binomial distribution is .. math:: P(N;n,p) = \frac{\Gamma(N+n)}{N!\Gamma(n)}p^{n}(1-p)^{N}, where :math:`n` is the number of successes, :math:`p` is the probability of success, :math:`N+n` is the number of trials, and :math:`\Gamma` is the gamma function. When :math:`n` is an integer, :math:`\frac{\Gamma(N+n)}{N!\Gamma(n)} = \binom{N+n-1}{N}`, which is the more common form of this term in the the pmf. The negative binomial distribution gives the probability of N failures given n successes, with a success on the last trial. If one throws a die repeatedly until the third time a "1" appears, then the probability distribution of the number of non-"1"s that appear before the third "1" is a negative binomial distribution. References ---------- .. [1] Weisstein, Eric W. "Negative Binomial Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/NegativeBinomialDistribution.html .. [2] Wikipedia, "Negative binomial distribution", https://en.wikipedia.org/wiki/Negative_binomial_distribution Examples -------- Draw samples from the distribution: A real world example. A company drills wild-cat oil exploration wells, each with an estimated probability of success of 0.1. What is the probability of having one success for each successive well, that is what is the probability of a single success after drilling 5 wells, after 6 wells, etc.? >>> s = np.random.default_rng().negative_binomial(1, 0.1, 100000) >>> for i in range(1, 11): # doctest: +SKIP ... probability = sum(s>> rng = np.random.default_rng() >>> n, p = 10, .5 # number of trials, probability of each trial >>> s = rng.binomial(n, p, 1000) # result of flipping a coin 10 times, tested 1000 times. A real world example. A company drills 9 wild-cat oil exploration wells, each with an estimated probability of success of 0.1. All nine wells fail. What is the probability of that happening? Let's do 20,000 trials of the model, and count the number that generate zero positive results. >>> sum(rng.binomial(9, 0.1, 20000) == 0)/20000. # answer = 0.38885, or 38%. triangular(left, mode, right, size=None) Draw samples from the triangular distribution over the interval ``[left, right]``. The triangular distribution is a continuous probability distribution with lower limit left, peak at mode, and upper limit right. Unlike the other distributions, these parameters directly define the shape of the pdf. Parameters ---------- left : float or array_like of floats Lower limit. mode : float or array_like of floats The value where the peak of the distribution occurs. The value must fulfill the condition ``left <= mode <= right``. right : float or array_like of floats Upper limit, must be larger than `left`. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``left``, ``mode``, and ``right`` are all scalars. Otherwise, ``np.broadcast(left, mode, right).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized triangular distribution. Notes ----- The probability density function for the triangular distribution is .. math:: P(x;l, m, r) = \begin{cases} \frac{2(x-l)}{(r-l)(m-l)}& \text{for $l \leq x \leq m$},\\ \frac{2(r-x)}{(r-l)(r-m)}& \text{for $m \leq x \leq r$},\\ 0& \text{otherwise}. \end{cases} The triangular distribution is often used in ill-defined problems where the underlying distribution is not known, but some knowledge of the limits and mode exists. Often it is used in simulations. References ---------- .. [1] Wikipedia, "Triangular distribution" https://en.wikipedia.org/wiki/Triangular_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> h = plt.hist(np.random.default_rng().triangular(-3, 0, 8, 100000), bins=200, ... density=True) >>> plt.show() wald(mean, scale, size=None) Draw samples from a Wald, or inverse Gaussian, distribution. As the scale approaches infinity, the distribution becomes more like a Gaussian. Some references claim that the Wald is an inverse Gaussian with mean equal to 1, but this is by no means universal. The inverse Gaussian distribution was first studied in relationship to Brownian motion. In 1956 M.C.K. Tweedie used the name inverse Gaussian because there is an inverse relationship between the time to cover a unit distance and distance covered in unit time. Parameters ---------- mean : float or array_like of floats Distribution mean, must be > 0. scale : float or array_like of floats Scale parameter, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(mean, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Wald distribution. Notes ----- The probability density function for the Wald distribution is .. math:: P(x;mean,scale) = \sqrt{\frac{scale}{2\pi x^3}}e^ \frac{-scale(x-mean)^2}{2\cdotp mean^2x} As noted above the inverse Gaussian distribution first arise from attempts to model Brownian motion. It is also a competitor to the Weibull for use in reliability modeling and modeling stock returns and interest rate processes. References ---------- .. [1] Brighton Webs Ltd., Wald Distribution, https://web.archive.org/web/20090423014010/http://www.brighton-webs.co.uk:80/distributions/wald.asp .. [2] Chhikara, Raj S., and Folks, J. Leroy, "The Inverse Gaussian Distribution: Theory : Methodology, and Applications", CRC Press, 1988. .. [3] Wikipedia, "Inverse Gaussian distribution" https://en.wikipedia.org/wiki/Inverse_Gaussian_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> h = plt.hist(np.random.default_rng().wald(3, 2, 100000), bins=200, density=True) >>> plt.show() rayleigh(scale=1.0, size=None) Draw samples from a Rayleigh distribution. The :math:`\chi` and Weibull distributions are generalizations of the Rayleigh. Parameters ---------- scale : float or array_like of floats, optional Scale, also equals the mode. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Rayleigh distribution. Notes ----- The probability density function for the Rayleigh distribution is .. math:: P(x;scale) = \frac{x}{scale^2}e^{\frac{-x^2}{2 \cdotp scale^2}} The Rayleigh distribution would arise, for example, if the East and North components of the wind velocity had identical zero-mean Gaussian distributions. Then the wind speed would have a Rayleigh distribution. References ---------- .. [1] Brighton Webs Ltd., "Rayleigh Distribution," https://web.archive.org/web/20090514091424/http://brighton-webs.co.uk:80/distributions/rayleigh.asp .. [2] Wikipedia, "Rayleigh distribution" https://en.wikipedia.org/wiki/Rayleigh_distribution Examples -------- Draw values from the distribution and plot the histogram >>> from matplotlib.pyplot import hist >>> rng = np.random.default_rng() >>> values = hist(rng.rayleigh(3, 100000), bins=200, density=True) Wave heights tend to follow a Rayleigh distribution. If the mean wave height is 1 meter, what fraction of waves are likely to be larger than 3 meters? >>> meanvalue = 1 >>> modevalue = np.sqrt(2 / np.pi) * meanvalue >>> s = rng.rayleigh(modevalue, 1000000) The percentage of waves larger than 3 meters is: >>> 100.*sum(s>3)/1000000. 0.087300000000000003 # random lognormal(mean=0.0, sigma=1.0, size=None) Draw samples from a log-normal distribution. Draw samples from a log-normal distribution with specified mean, standard deviation, and array shape. Note that the mean and standard deviation are not the values for the distribution itself, but of the underlying normal distribution it is derived from. Parameters ---------- mean : float or array_like of floats, optional Mean value of the underlying normal distribution. Default is 0. sigma : float or array_like of floats, optional Standard deviation of the underlying normal distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``sigma`` are both scalars. Otherwise, ``np.broadcast(mean, sigma).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized log-normal distribution. See Also -------- scipy.stats.lognorm : probability density function, distribution, cumulative density function, etc. Notes ----- A variable `x` has a log-normal distribution if `log(x)` is normally distributed. The probability density function for the log-normal distribution is: .. math:: p(x) = \frac{1}{\sigma x \sqrt{2\pi}} e^{(-\frac{(ln(x)-\mu)^2}{2\sigma^2})} where :math:`\mu` is the mean and :math:`\sigma` is the standard deviation of the normally distributed logarithm of the variable. A log-normal distribution results if a random variable is the *product* of a large number of independent, identically-distributed variables in the same way that a normal distribution results if the variable is the *sum* of a large number of independent, identically-distributed variables. References ---------- .. [1] Limpert, E., Stahel, W. A., and Abbt, M., "Log-normal Distributions across the Sciences: Keys and Clues," BioScience, Vol. 51, No. 5, May, 2001. https://stat.ethz.ch/~stahel/lognormal/bioscience.pdf .. [2] Reiss, R.D. and Thomas, M., "Statistical Analysis of Extreme Values," Basel: Birkhauser Verlag, 2001, pp. 31-32. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, sigma = 3., 1. # mean and standard deviation >>> s = rng.lognormal(mu, sigma, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 100, density=True, align='mid') >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, linewidth=2, color='r') >>> plt.axis('tight') >>> plt.show() Demonstrate that taking the products of random samples from a uniform distribution can be fit well by a log-normal probability density function. >>> # Generate a thousand samples: each is the product of 100 random >>> # values, drawn from a normal distribution. >>> rng = rng >>> b = [] >>> for i in range(1000): ... a = 10. + rng.standard_normal(100) ... b.append(np.product(a)) >>> b = np.array(b) / np.min(b) # scale values to be positive >>> count, bins, ignored = plt.hist(b, 100, density=True, align='mid') >>> sigma = np.std(np.log(b)) >>> mu = np.mean(np.log(b)) >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, color='r', linewidth=2) >>> plt.show() logistic(loc=0.0, scale=1.0, size=None) Draw samples from a logistic distribution. Samples are drawn from a logistic distribution with specified parameters, loc (location or mean, also median), and scale (>0). Parameters ---------- loc : float or array_like of floats, optional Parameter of the distribution. Default is 0. scale : float or array_like of floats, optional Parameter of the distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logistic distribution. See Also -------- scipy.stats.logistic : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Logistic distribution is .. math:: P(x) = P(x) = \frac{e^{-(x-\mu)/s}}{s(1+e^{-(x-\mu)/s})^2}, where :math:`\mu` = location and :math:`s` = scale. The Logistic distribution is used in Extreme Value problems where it can act as a mixture of Gumbel distributions, in Epidemiology, and by the World Chess Federation (FIDE) where it is used in the Elo ranking system, assuming the performance of each player is a logistically distributed random variable. References ---------- .. [1] Reiss, R.-D. and Thomas M. (2001), "Statistical Analysis of Extreme Values, from Insurance, Finance, Hydrology and Other Fields," Birkhauser Verlag, Basel, pp 132-133. .. [2] Weisstein, Eric W. "Logistic Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/LogisticDistribution.html .. [3] Wikipedia, "Logistic-distribution", https://en.wikipedia.org/wiki/Logistic_distribution Examples -------- Draw samples from the distribution: >>> loc, scale = 10, 1 >>> s = np.random.default_rng().logistic(loc, scale, 10000) >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, bins=50) # plot against distribution >>> def logist(x, loc, scale): ... return np.exp((loc-x)/scale)/(scale*(1+np.exp((loc-x)/scale))**2) >>> lgst_val = logist(bins, loc, scale) >>> plt.plot(bins, lgst_val * count.max() / lgst_val.max()) >>> plt.show() gumbel(loc=0.0, scale=1.0, size=None) Draw samples from a Gumbel distribution. Draw samples from a Gumbel distribution with specified location and scale. For more information on the Gumbel distribution, see Notes and References below. Parameters ---------- loc : float or array_like of floats, optional The location of the mode of the distribution. Default is 0. scale : float or array_like of floats, optional The scale parameter of the distribution. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Gumbel distribution. See Also -------- scipy.stats.gumbel_l scipy.stats.gumbel_r scipy.stats.genextreme weibull Notes ----- The Gumbel (or Smallest Extreme Value (SEV) or the Smallest Extreme Value Type I) distribution is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. The Gumbel is a special case of the Extreme Value Type I distribution for maximums from distributions with "exponential-like" tails. The probability density for the Gumbel distribution is .. math:: p(x) = \frac{e^{-(x - \mu)/ \beta}}{\beta} e^{ -e^{-(x - \mu)/ \beta}}, where :math:`\mu` is the mode, a location parameter, and :math:`\beta` is the scale parameter. The Gumbel (named for German mathematician Emil Julius Gumbel) was used very early in the hydrology literature, for modeling the occurrence of flood events. It is also used for modeling maximum wind speed and rainfall rates. It is a "fat-tailed" distribution - the probability of an event in the tail of the distribution is larger than if one used a Gaussian, hence the surprisingly frequent occurrence of 100-year floods. Floods were initially modeled as a Gaussian process, which underestimated the frequency of extreme events. It is one of a class of extreme value distributions, the Generalized Extreme Value (GEV) distributions, which also includes the Weibull and Frechet. The function has a mean of :math:`\mu + 0.57721\beta` and a variance of :math:`\frac{\pi^2}{6}\beta^2`. References ---------- .. [1] Gumbel, E. J., "Statistics of Extremes," New York: Columbia University Press, 1958. .. [2] Reiss, R.-D. and Thomas, M., "Statistical Analysis of Extreme Values from Insurance, Finance, Hydrology and Other Fields," Basel: Birkhauser Verlag, 2001. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, beta = 0, 0.1 # location and scale >>> s = rng.gumbel(mu, beta, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 30, density=True) >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp( -np.exp( -(bins - mu) /beta) ), ... linewidth=2, color='r') >>> plt.show() Show how an extreme value distribution can arise from a Gaussian process and compare to a Gaussian: >>> means = [] >>> maxima = [] >>> for i in range(0,1000) : ... a = rng.normal(mu, beta, 1000) ... means.append(a.mean()) ... maxima.append(a.max()) >>> count, bins, ignored = plt.hist(maxima, 30, density=True) >>> beta = np.std(maxima) * np.sqrt(6) / np.pi >>> mu = np.mean(maxima) - 0.57721*beta >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp(-np.exp(-(bins - mu)/beta)), ... linewidth=2, color='r') >>> plt.plot(bins, 1/(beta * np.sqrt(2 * np.pi)) ... * np.exp(-(bins - mu)**2 / (2 * beta**2)), ... linewidth=2, color='g') >>> plt.show() laplace(loc=0.0, scale=1.0, size=None) Draw samples from the Laplace or double exponential distribution with specified location (or mean) and scale (decay). The Laplace distribution is similar to the Gaussian/normal distribution, but is sharper at the peak and has fatter tails. It represents the difference between two independent, identically distributed exponential random variables. Parameters ---------- loc : float or array_like of floats, optional The position, :math:`\mu`, of the distribution peak. Default is 0. scale : float or array_like of floats, optional :math:`\lambda`, the exponential decay. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Laplace distribution. Notes ----- It has the probability density function .. math:: f(x; \mu, \lambda) = \frac{1}{2\lambda} \exp\left(-\frac{|x - \mu|}{\lambda}\right). The first law of Laplace, from 1774, states that the frequency of an error can be expressed as an exponential function of the absolute magnitude of the error, which leads to the Laplace distribution. For many problems in economics and health sciences, this distribution seems to model the data better than the standard Gaussian distribution. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] Kotz, Samuel, et. al. "The Laplace Distribution and Generalizations, " Birkhauser, 2001. .. [3] Weisstein, Eric W. "Laplace Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/LaplaceDistribution.html .. [4] Wikipedia, "Laplace distribution", https://en.wikipedia.org/wiki/Laplace_distribution Examples -------- Draw samples from the distribution >>> loc, scale = 0., 1. >>> s = np.random.default_rng().laplace(loc, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 30, density=True) >>> x = np.arange(-8., 8., .01) >>> pdf = np.exp(-abs(x-loc)/scale)/(2.*scale) >>> plt.plot(x, pdf) Plot Gaussian for comparison: >>> g = (1/(scale * np.sqrt(2 * np.pi)) * ... np.exp(-(x - loc)**2 / (2 * scale**2))) >>> plt.plot(x,g) power(a, size=None) Draws samples in [0, 1] from a power distribution with positive exponent a - 1. Also known as the power function distribution. Parameters ---------- a : float or array_like of floats Parameter of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized power distribution. Raises ------ ValueError If a < 1. Notes ----- The probability density function is .. math:: P(x; a) = ax^{a-1}, 0 \le x \le 1, a>0. The power function distribution is just the inverse of the Pareto distribution. It may also be seen as a special case of the Beta distribution. It is used, for example, in modeling the over-reporting of insurance claims. References ---------- .. [1] Christian Kleiber, Samuel Kotz, "Statistical size distributions in economics and actuarial sciences", Wiley, 2003. .. [2] Heckert, N. A. and Filliben, James J. "NIST Handbook 148: Dataplot Reference Manual, Volume 2: Let Subcommands and Library Functions", National Institute of Standards and Technology Handbook Series, June 2003. https://www.itl.nist.gov/div898/software/dataplot/refman2/auxillar/powpdf.pdf Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> samples = 1000 >>> s = rng.power(a, samples) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, bins=30) >>> x = np.linspace(0, 1, 100) >>> y = a*x**(a-1.) >>> normed_y = samples*np.diff(bins)[0]*y >>> plt.plot(x, normed_y) >>> plt.show() Compare the power function distribution to the inverse of the Pareto. >>> from scipy import stats # doctest: +SKIP >>> rvs = rng.power(5, 1000000) >>> rvsp = rng.pareto(5, 1000000) >>> xx = np.linspace(0,1,100) >>> powpdf = stats.powerlaw.pdf(xx,5) # doctest: +SKIP >>> plt.figure() >>> plt.hist(rvs, bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('power(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of 1 + Generator.pareto(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of stats.pareto(5)') weibull(a, size=None) Draw samples from a Weibull distribution. Draw samples from a 1-parameter Weibull distribution with the given shape parameter `a`. .. math:: X = (-ln(U))^{1/a} Here, U is drawn from the uniform distribution over (0,1]. The more common 2-parameter Weibull, including a scale parameter :math:`\lambda` is just :math:`X = \lambda(-ln(U))^{1/a}`. Parameters ---------- a : float or array_like of floats Shape parameter of the distribution. Must be nonnegative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Weibull distribution. See Also -------- scipy.stats.weibull_max scipy.stats.weibull_min scipy.stats.genextreme gumbel Notes ----- The Weibull (or Type III asymptotic extreme value distribution for smallest values, SEV Type III, or Rosin-Rammler distribution) is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. This class includes the Gumbel and Frechet distributions. The probability density for the Weibull distribution is .. math:: p(x) = \frac{a} {\lambda}(\frac{x}{\lambda})^{a-1}e^{-(x/\lambda)^a}, where :math:`a` is the shape and :math:`\lambda` the scale. The function has its peak (the mode) at :math:`\lambda(\frac{a-1}{a})^{1/a}`. When ``a = 1``, the Weibull distribution reduces to the exponential distribution. References ---------- .. [1] Waloddi Weibull, Royal Technical University, Stockholm, 1939 "A Statistical Theory Of The Strength Of Materials", Ingeniorsvetenskapsakademiens Handlingar Nr 151, 1939, Generalstabens Litografiska Anstalts Forlag, Stockholm. .. [2] Waloddi Weibull, "A Statistical Distribution Function of Wide Applicability", Journal Of Applied Mechanics ASME Paper 1951. .. [3] Wikipedia, "Weibull distribution", https://en.wikipedia.org/wiki/Weibull_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> s = rng.weibull(a, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> x = np.arange(1,100.)/50. >>> def weib(x,n,a): ... return (a / n) * (x / n)**(a - 1) * np.exp(-(x / n)**a) >>> count, bins, ignored = plt.hist(rng.weibull(5.,1000)) >>> x = np.arange(1,100.)/50. >>> scale = count.max()/weib(x, 1., 5.).max() >>> plt.plot(x, weib(x, 1., 5.)*scale) >>> plt.show() pareto(a, size=None) Draw samples from a Pareto II or Lomax distribution with specified shape. The Lomax or Pareto II distribution is a shifted Pareto distribution. The classical Pareto distribution can be obtained from the Lomax distribution by adding 1 and multiplying by the scale parameter ``m`` (see Notes). The smallest value of the Lomax distribution is zero while for the classical Pareto distribution it is ``mu``, where the standard Pareto distribution has location ``mu = 1``. Lomax can also be considered as a simplified version of the Generalized Pareto distribution (available in SciPy), with the scale set to one and the location set to zero. The Pareto distribution must be greater than zero, and is unbounded above. It is also known as the "80-20 rule". In this distribution, 80 percent of the weights are in the lowest 20 percent of the range, while the other 20 percent fill the remaining 80 percent of the range. Parameters ---------- a : float or array_like of floats Shape of the distribution. Must be positive. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Pareto distribution. See Also -------- scipy.stats.lomax : probability density function, distribution or cumulative density function, etc. scipy.stats.genpareto : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Pareto distribution is .. math:: p(x) = \frac{am^a}{x^{a+1}} where :math:`a` is the shape and :math:`m` the scale. The Pareto distribution, named after the Italian economist Vilfredo Pareto, is a power law probability distribution useful in many real world problems. Outside the field of economics it is generally referred to as the Bradford distribution. Pareto developed the distribution to describe the distribution of wealth in an economy. It has also found use in insurance, web page access statistics, oil field sizes, and many other problems, including the download frequency for projects in Sourceforge [1]_. It is one of the so-called "fat-tailed" distributions. References ---------- .. [1] Francis Hunt and Paul Johnson, On the Pareto Distribution of Sourceforge projects. .. [2] Pareto, V. (1896). Course of Political Economy. Lausanne. .. [3] Reiss, R.D., Thomas, M.(2001), Statistical Analysis of Extreme Values, Birkhauser Verlag, Basel, pp 23-30. .. [4] Wikipedia, "Pareto distribution", https://en.wikipedia.org/wiki/Pareto_distribution Examples -------- Draw samples from the distribution: >>> a, m = 3., 2. # shape and mode >>> s = (np.random.default_rng().pareto(a, 1000) + 1) * m Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 100, density=True) >>> fit = a*m**a / bins**(a+1) >>> plt.plot(bins, max(count)*fit/max(fit), linewidth=2, color='r') >>> plt.show() vonmises(mu, kappa, size=None) Draw samples from a von Mises distribution. Samples are drawn from a von Mises distribution with specified mode (mu) and dispersion (kappa), on the interval [-pi, pi]. The von Mises distribution (also known as the circular normal distribution) is a continuous probability distribution on the unit circle. It may be thought of as the circular analogue of the normal distribution. Parameters ---------- mu : float or array_like of floats Mode ("center") of the distribution. kappa : float or array_like of floats Dispersion of the distribution, has to be >=0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mu`` and ``kappa`` are both scalars. Otherwise, ``np.broadcast(mu, kappa).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized von Mises distribution. See Also -------- scipy.stats.vonmises : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability density for the von Mises distribution is .. math:: p(x) = \frac{e^{\kappa cos(x-\mu)}}{2\pi I_0(\kappa)}, where :math:`\mu` is the mode and :math:`\kappa` the dispersion, and :math:`I_0(\kappa)` is the modified Bessel function of order 0. The von Mises is named for Richard Edler von Mises, who was born in Austria-Hungary, in what is now the Ukraine. He fled to the United States in 1939 and became a professor at Harvard. He worked in probability theory, aerodynamics, fluid mechanics, and philosophy of science. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] von Mises, R., "Mathematical Theory of Probability and Statistics", New York: Academic Press, 1964. Examples -------- Draw samples from the distribution: >>> mu, kappa = 0.0, 4.0 # mean and dispersion >>> s = np.random.default_rng().vonmises(mu, kappa, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.special import i0 # doctest: +SKIP >>> plt.hist(s, 50, density=True) >>> x = np.linspace(-np.pi, np.pi, num=51) >>> y = np.exp(kappa*np.cos(x-mu))/(2*np.pi*i0(kappa)) # doctest: +SKIP >>> plt.plot(x, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() standard_t(df, size=None) Draw samples from a standard Student's t distribution with `df` degrees of freedom. A special case of the hyperbolic distribution. As `df` gets large, the result resembles that of the standard normal distribution (`standard_normal`). Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard Student's t distribution. Notes ----- The probability density function for the t distribution is .. math:: P(x, df) = \frac{\Gamma(\frac{df+1}{2})}{\sqrt{\pi df} \Gamma(\frac{df}{2})}\Bigl( 1+\frac{x^2}{df} \Bigr)^{-(df+1)/2} The t test is based on an assumption that the data come from a Normal distribution. The t test provides a way to test whether the sample mean (that is the mean calculated from the data) is a good estimate of the true mean. The derivation of the t-distribution was first published in 1908 by William Gosset while working for the Guinness Brewery in Dublin. Due to proprietary issues, he had to publish under a pseudonym, and so he used the name Student. References ---------- .. [1] Dalgaard, Peter, "Introductory Statistics With R", Springer, 2002. .. [2] Wikipedia, "Student's t-distribution" https://en.wikipedia.org/wiki/Student's_t-distribution Examples -------- From Dalgaard page 83 [1]_, suppose the daily energy intake for 11 women in kilojoules (kJ) is: >>> intake = np.array([5260., 5470, 5640, 6180, 6390, 6515, 6805, 7515, \ ... 7515, 8230, 8770]) Does their energy intake deviate systematically from the recommended value of 7725 kJ? We have 10 degrees of freedom, so is the sample mean within 95% of the recommended value? >>> s = np.random.default_rng().standard_t(10, size=100000) >>> np.mean(intake) 6753.636363636364 >>> intake.std(ddof=1) 1142.1232221373727 Calculate the t statistic, setting the ddof parameter to the unbiased value so the divisor in the standard deviation will be degrees of freedom, N-1. >>> t = (np.mean(intake)-7725)/(intake.std(ddof=1)/np.sqrt(len(intake))) >>> import matplotlib.pyplot as plt >>> h = plt.hist(s, bins=100, density=True) For a one-sided t-test, how far out in the distribution does the t statistic appear? >>> np.sum(s>> import matplotlib.pyplot as plt >>> s = np.random.default_rng().standard_cauchy(1000000) >>> s = s[(s>-25) & (s<25)] # truncate distribution so it plots well >>> plt.hist(s, bins=100) >>> plt.show() noncentral_chisquare(df, nonc, size=None) Draw samples from a noncentral chi-square distribution. The noncentral :math:`\chi^2` distribution is a generalization of the :math:`\chi^2` distribution. Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. .. versionchanged:: 1.10.0 Earlier NumPy versions required dfnum > 1. nonc : float or array_like of floats Non-centrality, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` and ``nonc`` are both scalars. Otherwise, ``np.broadcast(df, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral chi-square distribution. Notes ----- The probability density function for the noncentral Chi-square distribution is .. math:: P(x;df,nonc) = \sum^{\infty}_{i=0} \frac{e^{-nonc/2}(nonc/2)^{i}}{i!} P_{Y_{df+2i}}(x), where :math:`Y_{q}` is the Chi-square with q degrees of freedom. References ---------- .. [1] Wikipedia, "Noncentral chi-squared distribution" https://en.wikipedia.org/wiki/Noncentral_chi-squared_distribution Examples -------- Draw values from the distribution and plot the histogram >>> rng = np.random.default_rng() >>> import matplotlib.pyplot as plt >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() Draw values from a noncentral chisquare with very small noncentrality, and compare to a chisquare. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, .0000001, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> values2 = plt.hist(rng.chisquare(3, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> plt.plot(values[1][0:-1], values[0]-values2[0], 'ob') >>> plt.show() Demonstrate how large values of non-centrality lead to a more symmetric distribution. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() chisquare(df, size=None) Draw samples from a chi-square distribution. When `df` independent random variables, each with standard normal distributions (mean 0, variance 1), are squared and summed, the resulting distribution is chi-square (see Notes). This distribution is often used in hypothesis testing. Parameters ---------- df : float or array_like of floats Number of degrees of freedom, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized chi-square distribution. Raises ------ ValueError When `df` <= 0 or when an inappropriate `size` (e.g. ``size=-1``) is given. Notes ----- The variable obtained by summing the squares of `df` independent, standard normally distributed random variables: .. math:: Q = \sum_{i=0}^{\mathtt{df}} X^2_i is chi-square distributed, denoted .. math:: Q \sim \chi^2_k. The probability density function of the chi-squared distribution is .. math:: p(x) = \frac{(1/2)^{k/2}}{\Gamma(k/2)} x^{k/2 - 1} e^{-x/2}, where :math:`\Gamma` is the gamma function, .. math:: \Gamma(x) = \int_0^{-\infty} t^{x - 1} e^{-t} dt. References ---------- .. [1] NIST "Engineering Statistics Handbook" https://www.itl.nist.gov/div898/handbook/eda/section3/eda3666.htm Examples -------- >>> np.random.default_rng().chisquare(2,4) array([ 1.89920014, 9.00867716, 3.13710533, 5.62318272]) # random noncentral_f(dfnum, dfden, nonc, size=None) Draw samples from the noncentral F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters > 1. `nonc` is the non-centrality parameter. Parameters ---------- dfnum : float or array_like of floats Numerator degrees of freedom, must be > 0. .. versionchanged:: 1.14.0 Earlier NumPy versions required dfnum > 1. dfden : float or array_like of floats Denominator degrees of freedom, must be > 0. nonc : float or array_like of floats Non-centrality parameter, the sum of the squares of the numerator means, must be >= 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum``, ``dfden``, and ``nonc`` are all scalars. Otherwise, ``np.broadcast(dfnum, dfden, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral Fisher distribution. Notes ----- When calculating the power of an experiment (power = probability of rejecting the null hypothesis when a specific alternative is true) the non-central F statistic becomes important. When the null hypothesis is true, the F statistic follows a central F distribution. When the null hypothesis is not true, then it follows a non-central F statistic. References ---------- .. [1] Weisstein, Eric W. "Noncentral F-Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/NoncentralF-Distribution.html .. [2] Wikipedia, "Noncentral F-distribution", https://en.wikipedia.org/wiki/Noncentral_F-distribution Examples -------- In a study, testing for a specific alternative to the null hypothesis requires use of the Noncentral F distribution. We need to calculate the area in the tail of the distribution that exceeds the value of the F distribution for the null hypothesis. We'll plot the two probability distributions for comparison. >>> rng = np.random.default_rng() >>> dfnum = 3 # between group deg of freedom >>> dfden = 20 # within groups degrees of freedom >>> nonc = 3.0 >>> nc_vals = rng.noncentral_f(dfnum, dfden, nonc, 1000000) >>> NF = np.histogram(nc_vals, bins=50, density=True) >>> c_vals = rng.f(dfnum, dfden, 1000000) >>> F = np.histogram(c_vals, bins=50, density=True) >>> import matplotlib.pyplot as plt >>> plt.plot(F[1][1:], F[0]) >>> plt.plot(NF[1][1:], NF[0]) >>> plt.show() f(dfnum, dfden, size=None) Draw samples from an F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters must be greater than zero. The random variate of the F distribution (also known as the Fisher distribution) is a continuous probability distribution that arises in ANOVA tests, and is the ratio of two chi-square variates. Parameters ---------- dfnum : float or array_like of floats Degrees of freedom in numerator, must be > 0. dfden : float or array_like of float Degrees of freedom in denominator, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum`` and ``dfden`` are both scalars. Otherwise, ``np.broadcast(dfnum, dfden).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Fisher distribution. See Also -------- scipy.stats.f : probability density function, distribution or cumulative density function, etc. Notes ----- The F statistic is used to compare in-group variances to between-group variances. Calculating the distribution depends on the sampling, and so it is a function of the respective degrees of freedom in the problem. The variable `dfnum` is the number of samples minus one, the between-groups degrees of freedom, while `dfden` is the within-groups degrees of freedom, the sum of the number of samples in each group minus the number of groups. References ---------- .. [1] Glantz, Stanton A. "Primer of Biostatistics.", McGraw-Hill, Fifth Edition, 2002. .. [2] Wikipedia, "F-distribution", https://en.wikipedia.org/wiki/F-distribution Examples -------- An example from Glantz[1], pp 47-40: Two groups, children of diabetics (25 people) and children from people without diabetes (25 controls). Fasting blood glucose was measured, case group had a mean value of 86.1, controls had a mean value of 82.2. Standard deviations were 2.09 and 2.49 respectively. Are these data consistent with the null hypothesis that the parents diabetic status does not affect their children's blood glucose levels? Calculating the F statistic from the data gives a value of 36.01. Draw samples from the distribution: >>> dfnum = 1. # between group degrees of freedom >>> dfden = 48. # within groups degrees of freedom >>> s = np.random.default_rng().f(dfnum, dfden, 1000) The lower bound for the top 1% of the samples is : >>> np.sort(s)[-10] 7.61988120985 # random So there is about a 1% chance that the F statistic will exceed 7.62, the measured value is 36, so the null hypothesis is rejected at the 1% level. gamma(shape, scale=1.0, size=None) Draw samples from a Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, `shape` (sometimes designated "k") and `scale` (sometimes designated "theta"), where both parameters are > 0. Parameters ---------- shape : float or array_like of floats The shape of the gamma distribution. Must be non-negative. scale : float or array_like of floats, optional The scale of the gamma distribution. Must be non-negative. Default is equal to 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(shape, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 2. # mean=4, std=2*sqrt(2) >>> s = np.random.default_rng().gamma(shape, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, ignored = plt.hist(s, 50, density=True) >>> y = bins**(shape-1)*(np.exp(-bins/scale) / # doctest: +SKIP ... (sps.gamma(shape)*scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() standard_gamma(shape, size=None, dtype='d', out=None) Draw samples from a standard Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, shape (sometimes designated "k") and scale=1. Parameters ---------- shape : float or array_like of floats Parameter, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` is a scalar. Otherwise, ``np.array(shape).size`` samples are drawn. dtype : {str, dtype}, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 1. # mean and width >>> s = np.random.default_rng().standard_gamma(shape, 1000000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, ignored = plt.hist(s, 50, density=True) >>> y = bins**(shape-1) * ((np.exp(-bins/scale))/ # doctest: +SKIP ... (sps.gamma(shape) * scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() normal(loc=0.0, scale=1.0, size=None) Draw random samples from a normal (Gaussian) distribution. The probability density function of the normal distribution, first derived by De Moivre and 200 years later by both Gauss and Laplace independently [2]_, is often called the bell curve because of its characteristic shape (see the example below). The normal distributions occurs often in nature. For example, it describes the commonly occurring distribution of samples influenced by a large number of tiny, random disturbances, each with its own unique distribution [2]_. Parameters ---------- loc : float or array_like of floats Mean ("centre") of the distribution. scale : float or array_like of floats Standard deviation (spread or "width") of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized normal distribution. See Also -------- scipy.stats.norm : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gaussian distribution is .. math:: p(x) = \frac{1}{\sqrt{ 2 \pi \sigma^2 }} e^{ - \frac{ (x - \mu)^2 } {2 \sigma^2} }, where :math:`\mu` is the mean and :math:`\sigma` the standard deviation. The square of the standard deviation, :math:`\sigma^2`, is called the variance. The function has its peak at the mean, and its "spread" increases with the standard deviation (the function reaches 0.607 times its maximum at :math:`x + \sigma` and :math:`x - \sigma` [2]_). This implies that :meth:`normal` is more likely to return samples lying close to the mean, rather than those far away. References ---------- .. [1] Wikipedia, "Normal distribution", https://en.wikipedia.org/wiki/Normal_distribution .. [2] P. R. Peebles Jr., "Central Limit Theorem" in "Probability, Random Variables and Random Signal Principles", 4th ed., 2001, pp. 51, 51, 125. Examples -------- Draw samples from the distribution: >>> mu, sigma = 0, 0.1 # mean and standard deviation >>> s = np.random.default_rng().normal(mu, sigma, 1000) Verify the mean and the variance: >>> abs(mu - np.mean(s)) 0.0 # may vary >>> abs(sigma - np.std(s, ddof=1)) 0.1 # may vary Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 30, density=True) >>> plt.plot(bins, 1/(sigma * np.sqrt(2 * np.pi)) * ... np.exp( - (bins - mu)**2 / (2 * sigma**2) ), ... linewidth=2, color='r') >>> plt.show() Two-by-four array of samples from N(3, 6.25): >>> np.random.default_rng().normal(3, 2.5, size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random standard_normal(size=None, dtype='d', out=None) Draw samples from a standard Normal distribution (mean=0, stdev=1). Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : {str, dtype}, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray A floating-point array of shape ``size`` of drawn samples, or a single sample if ``size`` was not specified. See Also -------- normal : Equivalent function with additional ``loc`` and ``scale`` arguments for setting the mean and standard deviation. Notes ----- For random samples from :math:`N(\mu, \sigma^2)`, use one of:: mu + sigma * gen.standard_normal(size=...) gen.normal(mu, sigma, size=...) Examples -------- >>> rng = np.random.default_rng() >>> rng.standard_normal() 2.1923875335537315 #random >>> s = rng.standard_normal(8000) >>> s array([ 0.6888893 , 0.78096262, -0.89086505, ..., 0.49876311, # random -0.38672696, -0.4685006 ]) # random >>> s.shape (8000,) >>> s = rng.standard_normal(size=(3, 4, 2)) >>> s.shape (3, 4, 2) Two-by-four array of samples from :math:`N(3, 6.25)`: >>> 3 + 2.5 * rng.standard_normal(size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random uniform(low=0.0, high=1.0, size=None) Draw samples from a uniform distribution. Samples are uniformly distributed over the half-open interval ``[low, high)`` (includes low, but excludes high). In other words, any value within the given interval is equally likely to be drawn by `uniform`. Parameters ---------- low : float or array_like of floats, optional Lower boundary of the output interval. All values generated will be greater than or equal to low. The default value is 0. high : float or array_like of floats Upper boundary of the output interval. All values generated will be less than high. The default value is 1.0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``low`` and ``high`` are both scalars. Otherwise, ``np.broadcast(low, high).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized uniform distribution. See Also -------- integers : Discrete uniform distribution, yielding integers. random : Floats uniformly distributed over ``[0, 1)``. Notes ----- The probability density function of the uniform distribution is .. math:: p(x) = \frac{1}{b - a} anywhere within the interval ``[a, b)``, and zero elsewhere. When ``high`` == ``low``, values of ``low`` will be returned. If ``high`` < ``low``, the results are officially undefined and may eventually raise an error, i.e. do not rely on this function to behave when passed arguments satisfying that inequality condition. Examples -------- Draw samples from the distribution: >>> s = np.random.default_rng().uniform(-1,0,1000) All values are within the given interval: >>> np.all(s >= -1) True >>> np.all(s < 0) True Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, ignored = plt.hist(s, 15, density=True) >>> plt.plot(bins, np.ones_like(bins), linewidth=2, color='r') >>> plt.show() choice(a, size=None, replace=True, p=None, axis=0): Generates a random sample from a given 1-D array Parameters ---------- a : 1-D array-like or int If an ndarray, a random sample is generated from its elements. If an int, the random sample is generated as if a were np.arange(a) size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn from the 1-d `a`. If `a` has more than one dimension, the `size` shape will be inserted into the `axis` dimension, so the output ``ndim`` will be ``a.ndim - 1 + len(size)``. Default is None, in which case a single value is returned. replace : boolean, optional Whether the sample is with or without replacement p : 1-D array-like, optional The probabilities associated with each entry in a. If not given the sample assumes a uniform distribution over all entries in a. axis : int, optional The axis along which the selection is performed. The default, 0, selects by row. shuffle : boolean, optional Whether the sample is shuffled when sampling without replacement. Default is True, False provides a speedup. Returns ------- samples : single item or ndarray The generated random samples Raises ------ ValueError If a is an int and less than zero, if p is not 1-dimensional, if a is array-like with a size 0, if p is not a vector of probabilities, if a and p have different lengths, or if replace=False and the sample size is greater than the population size. See Also -------- integers, shuffle, permutation Examples -------- Generate a uniform random sample from np.arange(5) of size 3: >>> rng = np.random.default_rng() >>> rng.choice(5, 3) array([0, 3, 4]) # random >>> #This is equivalent to rng.integers(0,5,3) Generate a non-uniform random sample from np.arange(5) of size 3: >>> rng.choice(5, 3, p=[0.1, 0, 0.3, 0.6, 0]) array([3, 3, 0]) # random Generate a uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False) array([3,1,0]) # random >>> #This is equivalent to rng.permutation(np.arange(5))[:3] Generate a non-uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False, p=[0.1, 0, 0.3, 0.6, 0]) array([2, 3, 0]) # random Any of the above can be repeated with an arbitrary array-like instead of just integers. For instance: >>> aa_milne_arr = ['pooh', 'rabbit', 'piglet', 'Christopher'] >>> rng.choice(aa_milne_arr, 5, p=[0.5, 0.1, 0.1, 0.3]) array(['pooh', 'pooh', 'pooh', 'Christopher', 'piglet'], # random dtype='>> np.random.default_rng().bytes(10) ' eh\x85\x022SZ\xbf\xa4' #random integers(low, high=None, size=None, dtype='int64', endpoint=False) Return random integers from `low` (inclusive) to `high` (exclusive), or if endpoint=True, `low` (inclusive) to `high` (inclusive). Replaces `RandomState.randint` (with endpoint=False) and `RandomState.random_integers` (with endpoint=True) Return random integers from the "discrete uniform" distribution of the specified dtype. If `high` is None (the default), then results are from 0 to `low`. Parameters ---------- low : int or array-like of ints Lowest (signed) integers to be drawn from the distribution (unless ``high=None``, in which case this parameter is 0 and this value is used for `high`). high : int or array-like of ints, optional If provided, one above the largest (signed) integer to be drawn from the distribution (see above for behavior if ``high=None``). If array-like, must contain integer values size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : {str, dtype}, optional Desired dtype of the result. All dtypes are determined by their name, i.e., 'int64', 'int', etc, so byteorder is not available and a specific precision may have different C types depending on the platform. The default value is `np.int_`. endpoint : bool, optional If true, sample from the interval [low, high] instead of the default [low, high) Defaults to False Returns ------- out : int or ndarray of ints `size`-shaped array of random integers from the appropriate distribution, or a single such random int if `size` not provided. Notes ----- When using broadcasting with uint64 dtypes, the maximum value (2**64) cannot be represented as a standard integer type. The high array (or low if high is None) must have object dtype, e.g., array([2**64]). Examples -------- >>> rng = np.random.default_rng() >>> rng.integers(2, size=10) array([1, 0, 0, 0, 1, 1, 0, 0, 1, 0]) # random >>> rng.integers(1, size=10) array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0]) Generate a 2 x 4 array of ints between 0 and 4, inclusive: >>> rng.integers(5, size=(2, 4)) array([[4, 0, 2, 1], [3, 2, 2, 0]]) # random Generate a 1 x 3 array with 3 different upper bounds >>> rng.integers(1, [3, 5, 10]) array([2, 2, 9]) # random Generate a 1 by 3 array with 3 different lower bounds >>> rng.integers([1, 5, 7], 10) array([9, 8, 7]) # random Generate a 2 by 4 array using broadcasting with dtype of uint8 >>> rng.integers([1, 3, 5, 7], [[10], [20]], dtype=np.uint8) array([[ 8, 6, 9, 7], [ 1, 16, 9, 12]], dtype=uint8) # random References ---------- .. [1] Daniel Lemire., "Fast Random Integer Generation in an Interval", ACM Transactions on Modeling and Computer Simulation 29 (1), 2019, http://arxiv.org/abs/1805.10941. standard_exponential(size=None, dtype='d', method='zig', out=None) Draw samples from the standard exponential distribution. `standard_exponential` is identical to the exponential distribution with a scale parameter of 1. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. method : str, optional Either 'inv' or 'zig'. 'inv' uses the default inverse CDF method. 'zig' uses the much faster Ziggurat method of Marsaglia and Tsang. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray Drawn samples. Examples -------- Output a 3x8000 array: >>> n = np.random.default_rng().standard_exponential((3, 8000)) exponential(scale=1.0, size=None) Draw samples from an exponential distribution. Its probability density function is .. math:: f(x; \frac{1}{\beta}) = \frac{1}{\beta} \exp(-\frac{x}{\beta}), for ``x > 0`` and 0 elsewhere. :math:`\beta` is the scale parameter, which is the inverse of the rate parameter :math:`\lambda = 1/\beta`. The rate parameter is an alternative, widely used parameterization of the exponential distribution [3]_. The exponential distribution is a continuous analogue of the geometric distribution. It describes many common situations, such as the size of raindrops measured over many rainstorms [1]_, or the time between page requests to Wikipedia [2]_. Parameters ---------- scale : float or array_like of floats The scale parameter, :math:`\beta = 1/\lambda`. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized exponential distribution. References ---------- .. [1] Peyton Z. Peebles Jr., "Probability, Random Variables and Random Signal Principles", 4th ed, 2001, p. 57. .. [2] Wikipedia, "Poisson process", https://en.wikipedia.org/wiki/Poisson_process .. [3] Wikipedia, "Exponential distribution", https://en.wikipedia.org/wiki/Exponential_distribution beta(a, b, size=None) Draw samples from a Beta distribution. The Beta distribution is a special case of the Dirichlet distribution, and is related to the Gamma distribution. It has the probability distribution function .. math:: f(x; a,b) = \frac{1}{B(\alpha, \beta)} x^{\alpha - 1} (1 - x)^{\beta - 1}, where the normalization, B, is the beta function, .. math:: B(\alpha, \beta) = \int_0^1 t^{\alpha - 1} (1 - t)^{\beta - 1} dt. It is often seen in Bayesian inference and order statistics. Parameters ---------- a : float or array_like of floats Alpha, positive (>0). b : float or array_like of floats Beta, positive (>0). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` and ``b`` are both scalars. Otherwise, ``np.broadcast(a, b).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized beta distribution. random(size=None, dtype='d', out=None) Return random floats in the half-open interval [0.0, 1.0). Results are from the "continuous uniform" distribution over the stated interval. To sample :math:`Unif[a, b), b > a` multiply the output of `random` by `(b-a)` and add `a`:: (b - a) * random() + a Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : {str, dtype}, optional Desired dtype of the result, either 'd' (or 'float64') or 'f' (or 'float32'). All dtypes are determined by their name. The default value is 'd'. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray of floats Array of random floats of shape `size` (unless ``size=None``, in which case a single float is returned). Examples -------- >>> rng = np.random.default_rng() >>> rng.random() 0.47108547995356098 # random >>> type(rng.random()) >>> rng.random((5,)) array([ 0.30220482, 0.86820401, 0.1654503 , 0.11659149, 0.54323428]) # random Three-by-two array of random numbers from [-5, 0): >>> 5 * rng.random((3, 2)) - 5 array([[-3.99149989, -0.52338984], # random [-2.99091858, -0.79479508], [-1.23204345, -1.75224494]]) 0,!,p,%,p$,0,,0,0,!,,p,,,0,!,!,%,0,@,`,$,p,,,0,0,!,p,,,0,,0,!,p,,,0,@",P",0,@",P",,0,`",0,`",,0,0,`",0,,p,0,%,0,%,0,%,0,,,0,,,0,,,0,@,@,0,,0,@,,0,@,,,0,p,`,0,p,`,0,`,0,%,0,`,0,,0,0,0,`,0,@,",0,#,`,,p,,0,0#,0,0,, %,0,,$,,$,@$,,, ,,0%,@,,!,!,,,P, ,P+, @+,30+,D$ +,C7+,C&+,@C=*, *,BA*, *,B#*,`B&*, B<*,7#*,  p*, `*, P*,@ @*, 0*,   *, *,` *, ), ),A%), ), ), ), ), ),`"p),A2`),`A*P), A(@),@+0),@# ), ),` ), "(, (,@ (,  (, (,` (,`@&(, @*(,?%p(,?%`(,@ !P(, !@(,@ 0(,`  (, (,@ (, ',L ',`?2', ', ?"',>Z',>@',@>0', p',@ `', P', @',>$0',`  ',$ ',=)', &, &,4 &, =&,  &, &,a &," &, p&,<!`&,<$P&, "@&,`<00&,@7+ &,7*&, &,0 &,7%,<F%, %, %, %,  %,  %,  %,0%,;&p%,`5`%, P%, @%, 0%,   %, %,c %, $, $, $,[ $,  $, $,P $, $, p$, h$, `$, P$,@@$, 0$,  $, $,;)$,` $,s#, #, #,H #,H #,( #,@;4#,#, x p#, `#, P#, @#, 0#,  #,:T#, #, ",` ", ", ", ", ",:%",`:)", p", `", X", P", H", @", 8", 0",  ",  ", ", !, !, !, !, !, !, !, !, p!, `!,  P!, @!, 8!, 0!,  !, !, !,  ,@  ,8  ,8  ,  ,  ,0  ,0  , p , ` ,P P , @ ,J0 ,  , :6 ,(  ,@t), ,a,  , , , ,~ ,x ,r p,r `,z P,z @,  0,  ,T ,@ ,u , ,l , , , ,p , , p,f h,f `, X, P, @,k 0,  , ,f , , , ,a , ,Hd ,7,, ,\ , p,S `, P,@ @,N 8,N 0,9# , , , ,9',`92,I , , , ,@7 , ,p, `, P,@ @, 0,;  ,; ,6 ,7 , 9$, ,` ,` , ,83,1 ,1 , p,= `,P,@ @, 0,  , , , , , ,Z , ,',8",t ,@ , ,x p, `, P, @,  0,,  ,,S ,, ,q ,, , ,`8#, , ,' ,M p,M `, P, @, 0,g  , ,f , ,X ,G , ,<,A ,; ,` , p,[ `,p P, @, 0,  , ,R ,5 ,5 ," ,/ ,/ , , ,E , ,) x,) p, `,@P, @,# 8,# 0,  , ,  ,,`, , ,,z , , , p, `, P, @,@!0,   , , , ,< ,8N, , , ,3 , , p, `, P,@qY @, 0,  , , , , , ,p ,7&,`g ,| , g%,u ,n p, [ `,D P, @, 0,*  , N , ,  , ,@D  ,+@"ADp"G`d ++ n@s "7#`+0+ +D<K`++x.e.p:0Zn@@~}`}|zPs@ :@D.C+@ 1s pj +++++V)p.P@p.P;(0Hs "";$D".PH++`p1@)@p><IVP@+ +<@+30y+P~+0J|+pp+ f+L`]+ N+$plB+-p7+*+p@+@ +` +`+***`*0*p** *0 *|*0q*0@h*, Z*~pM*v0E*q ;*PЦ4*KP "*g *q*@)>A) )p`)@)Z HP)UGCC: (GNU) 4.1.2 20080704 (Red Hat 4.1.2-55)GCC: (GNU) 4.8.2 20140120 (Red Hat 4.8.2-15)GCC: (GNU) 4.1.2 20080704 (Red Hat 4.1.2-55).shstrtab.note.gnu.build-id.gnu.hash.dynsym.dynstr.gnu.version.gnu.version_r.rela.dyn.rela.plt.init.text.fini.rodata.eh_frame_hdr.eh_frame.init_array.fini_array.jcr.data.rel.ro.dynamic.got.got.plt.data.bss.comment $o( HH(0pp8o22.EoH4H4@T44w^B@@h88cPP` n't z} m m v v \C) ) ) )  ) ) ) `)` 5 , (/  q