{
  "metadata": {
    "version": 1,
    "disable_limits": false
  },
  "nodes": [
    {
      "node_id": "8ceb5424-4b13-4ca8-a498-5bc084d15b17",
      "type": "SOURCE",
      "operator": "sagemaker.s3_source_0.1",
      "parameters": {
        "dataset_definition": {
          "__typename": "S3CreateDatasetDefinitionOutput",
          "datasetSourceType": "S3",
          "name": "signup_outcomes.csv",
          "description": null,
          "s3ExecutionContext": {
            "__typename": "S3ExecutionContext",
            "s3Uri": "s3://${bucket}/${prefix}/data/raw/signup_outcomes.csv",
            "s3ContentType": "csv",
            "s3HasHeader": true
          }
        }
      },
      "inputs": [],
      "outputs": [
        {
          "name": "default",
          "sampling": {
            "sampling_method": "sample_by_limit",
            "limit_rows": 50000
          }
        }
      ]
    },
    {
      "node_id": "c679c92d-ae35-440c-9dc2-649164010bd2",
      "type": "TRANSFORM",
      "operator": "sagemaker.spark.infer_and_cast_type_0.1",
      "parameters": {},
      "trained_parameters": {
        "schema": {
          "ip_address": "string",
          "email_address": "string",
          "EVENT_LABEL": "string"
        }
      },
      "inputs": [
        {
          "name": "default",
          "node_id": "8ceb5424-4b13-4ca8-a498-5bc084d15b17",
          "output_name": "default"
        }
      ],
      "outputs": [
        {
          "name": "default"
        }
      ]
    },
    {
      "node_id": "7c154ee7-e3dd-4fae-a3ab-29c60900de5b",
      "type": "TRANSFORM",
      "operator": "sagemaker.spark.handle_missing_0.1",
      "parameters": {
        "operator": "Fill missing",
        "fill_missing_parameters": {
          "input_column": "ip_address",
          "fill_value": "0.0.0.0"
        },
        "impute_parameters": {
          "column_type": "Numeric",
          "numeric_parameters": {
            "strategy": "Approximate Median"
          }
        }
      },
      "inputs": [
        {
          "name": "df",
          "node_id": "c679c92d-ae35-440c-9dc2-649164010bd2",
          "output_name": "default"
        }
      ],
      "outputs": [
        {
          "name": "default"
        }
      ]
    },
    {
      "node_id": "9268ca24-7772-4b79-bf42-e11ec0bb7c90",
      "type": "TRANSFORM",
      "operator": "sagemaker.spark.handle_missing_0.1",
      "parameters": {
        "operator": "Fill missing",
        "fill_missing_parameters": {
          "input_column": "email_address",
          "fill_value": "synth_missing@email.com"
        },
        "impute_parameters": {
          "column_type": "Numeric",
          "numeric_parameters": {
            "strategy": "Approximate Median"
          }
        }
      },
      "inputs": [
        {
          "name": "df",
          "node_id": "7c154ee7-e3dd-4fae-a3ab-29c60900de5b",
          "output_name": "default"
        }
      ],
      "outputs": [
        {
          "name": "default"
        }
      ]
    },
    {
      "node_id": "d8c363ff-9f7c-4b44-a7dc-44004df88c29",
      "type": "TRANSFORM",
      "operator": "sagemaker.spark.custom_pandas_0.1",
      "parameters": {
        "code": "# Table is available as variable `df`\nimport pandas as pd\ndf['EventTime'] = pd.to_datetime('now').timestamp()"
      },
      "inputs": [
        {
          "name": "df",
          "node_id": "9268ca24-7772-4b79-bf42-e11ec0bb7c90",
          "output_name": "default"
        }
      ],
      "outputs": [
        {
          "name": "default"
        }
      ]
    }
  ]
}