{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Use Case 1"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Find the companies where X has worked, and their roles at those companies"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## What questions would we have to ask of our data?"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"_\"Which companies has X worked for, and in what roles?\"_\n",
"\n",
"Reviewing this question we can identify several entities, attributes and relationships. We have the concept of a _company_, a _person_ (X), and a _role_. Further, a person _worked for_ a company.\n",
"\n",
"_Company_ and _Person_ are both entities, which we'll model as vertices with appropriate labels. For now, we'll assume a direct relationship between a person and a company: a person _WORKED FOR_ a company. We'll make _role_ an attribute of this relationship.\n",
"\n",
"Adding in a few properties – _firstName_ and _lastName_ for a person, _name_ for a company – we end up with the following data model:"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"
"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Keep it simple"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Over the course of this exercise we'll see _role_ change place several times. At this stage it's a simple attribute of a relationship. In later steps we'll see it promoted to being a vertex in its own right.\n",
"\n",
"As far as our current use case is concerned, role appears to be a simple value type, much like colour, height or weight. If it were a complex value type with several fields – such as address – or if there were some explicit structural relations between values – as there are in a category hierarchy – we would consider making it a vertex from the outset."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Sample dataset"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We'll now create a sample dataset in line with our model. We'll include enough data to ensure that our queries have to exclude some portions of the graph in order to return a correct result."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"
"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Creating our sample data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%load_ext ipython_unittest\n",
"%run '../util/neptune.py'"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"neptune.clear()\n",
"g = neptune.graphTraversal()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"(g.\n",
" addV('Person').property(id,'p-1').property('firstName','Martha').property('lastName','Rivera').\n",
" addV('Person').property(id,'p-2').property('firstName','Richard').property('lastName','Roe').\n",
" addV('Person').property(id,'p-3').property('firstName','Li').property('lastName','Juan').\n",
" addV('Person').property(id,'p-4').property('firstName','John').property('lastName','Stiles').\n",
" addV('Person').property(id,'p-5').property('firstName','Saanvi').property('lastName','Sarkar').\n",
" addV('Company').property(id,'c-1').property('name','Example Corp').\n",
" addV('Company').property(id,'c-2').property('name','AnyCompany').\n",
" V('p-1').addE('WORKED_FOR').to(V('c-1')).property('role','Principal Analyst'). \n",
" V('p-2').addE('WORKED_FOR').to(V('c-1')).property('role','Senior Analyst'). \n",
" V('p-3').addE('WORKED_FOR').to(V('c-1')).property('role','Analyst').\n",
" V('p-4').addE('WORKED_FOR').to(V('c-1')).property('role','Analyst'). \n",
" V('p-5').addE('WORKED_FOR').to(V('c-2')).property('role','Manager').\n",
" V('p-3').addE('WORKED_FOR').to(V('c-2')).property('role','Associate Analyst').\n",
" toList())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Querying the data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Query 1 – Which companies has Li worked for, and in what roles?"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To answer this question, we'll have to perform the following steps:\n",
"\n",
" 1. Start at the Person vertex representing Li\n",
" 2. Follow WORKED_FOR edges to find each Company for whom Li has worked\n",
" 3. Select the Company details, and the role property of the relationship"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Write a failing unit test"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%unittest\n",
"\n",
"results = None # TODO\n",
"\n",
"assert results == [{'company': 'Example Corp', 'role': 'Analyst'}, \n",
" {'company': 'AnyCompany', 'role': 'Associate Analyst'}]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### And write the query to make it pass"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%unittest\n",
"\n",
"results = (g.V('p-3').\n",
" outE('WORKED_FOR').as_('e').\n",
" otherV().\n",
" project('company', 'role').\n",
" by('name').\n",
" by(select('e').values('role')).\n",
" toList())\n",
"\n",
"assert results == [{'company': 'Example Corp', 'role': 'Analyst'}, \n",
" {'company': 'AnyCompany', 'role': 'Associate Analyst'}]"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "conda_python3",
"language": "python",
"name": "conda_python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.5"
}
},
"nbformat": 4,
"nbformat_minor": 2
}