{ "metadata": { "version": 1, "disable_limits": false }, "nodes": [ { "node_id": "2704b888-59de-413c-bc15-8ac20cde8375", "type": "SOURCE", "operator": "sagemaker.s3_source_0.1", "parameters": { "dataset_definition": { "__typename": "S3CreateDatasetDefinitionOutput", "datasetSourceType": "S3", "name": "dataset.csv", "description": null, "s3ExecutionContext": { "__typename": "S3ExecutionContext", "s3Uri": "s3://sagemaker-us-east-1-631450739534/sagemaker/DEMO-xgboost-customer-churn-connect/data/dataset.csv", "s3ContentType": "csv", "s3HasHeader": true, "s3FieldDelimiter": ",", "s3DirIncludesNested": false, "s3AddsFilenameColumn": false } } }, "inputs": [], "outputs": [ { "name": "default" } ] }, { "node_id": "99d1a36d-66f8-4969-80f0-1f0770c58bf2", "type": "TRANSFORM", "operator": "sagemaker.spark.infer_and_cast_type_0.1", "parameters": {}, "trained_parameters": { "schema": { "State": "string", "Account_Length": "long", "Area_Code": "long", "customerID": "long", "Int_l_Plan": "string", "VMail_Plan": "string", "VMail_Message": "long", "Day_Mins": "float", "Day_Calls": "long", "Day_Charge": "float", "Eve_Mins": "float", "Eve_Calls": "long", "Eve_Charge": "float", "Night_Mins": "float", "Night_Calls": "long", "Night_Charge": "float", "Intl_Mins": "float", "Intl_Calls": "long", "Intl_Charge": "float", "CustServ_Calls": "long", "Churn": "bool", "pastSenti_nut": "long", "pastSenti_pos": "long", "pastSenti_neg": "long", "mth_remain": "long" } }, "inputs": [ { "name": "default", "node_id": "2704b888-59de-413c-bc15-8ac20cde8375", "output_name": "default" } ], "outputs": [ { "name": "default" } ] }, { "node_id": "0ef7c84a-b3d9-493a-b9ae-c52f576f2bfe", "type": "TRANSFORM", "operator": "sagemaker.spark.custom_pandas_0.1", "parameters": { "code": "# Table is available as variable `df`\nimport pandas as pd\n\ndf = df.drop([\"Day_Charge\", \"Eve_Charge\", \"Night_Charge\", \"Intl_Charge\", \"Area_Code\",\"State\"], axis=1)\ndf = pd.get_dummies(df)\n\n# Add event time for features store\ndf['event_time']=pd.to_datetime('now').timestamp()\n\n" }, "inputs": [ { "name": "df", "node_id": "881fb0fb-d960-4b93-a2ef-39ac23255e2f", "output_name": "default" } ], "outputs": [ { "name": "default" } ] }, { "node_id": "881fb0fb-d960-4b93-a2ef-39ac23255e2f", "type": "TRANSFORM", "operator": "sagemaker.spark.cast_single_data_type_0.1", "parameters": { "column": "Churn", "original_data_type": "Boolean", "data_type": "string" }, "inputs": [ { "name": "df", "node_id": "99d1a36d-66f8-4969-80f0-1f0770c58bf2", "output_name": "default" } ], "outputs": [ { "name": "default" } ] }, { "node_id": "2db072bc-fcd4-4793-bdf9-f03d5c0c2c49", "type": "TRANSFORM", "operator": "sagemaker.spark.manage_columns_0.1", "parameters": { "operator": "Drop column", "drop_column_parameters": { "column_to_drop": "Churn_false" } }, "inputs": [ { "name": "df", "node_id": "0ef7c84a-b3d9-493a-b9ae-c52f576f2bfe", "output_name": "default" } ], "outputs": [ { "name": "default" } ] }, { "node_id": "c71b352f-7bf5-4570-a80e-9acae0b6bf3f", "type": "TRANSFORM", "operator": "sagemaker.spark.manage_columns_0.1", "parameters": { "operator": "Move column", "move_column_parameters": { "move_type": "Move to start", "move_to_start_parameters": { "column_to_move": "Churn_true" } }, "drop_column_parameters": {} }, "inputs": [ { "name": "df", "node_id": "2db072bc-fcd4-4793-bdf9-f03d5c0c2c49", "output_name": "default" } ], "outputs": [ { "name": "default" } ] } ] }