Sample Post Body
The following sample represents the post body of a Create Job Training request pointing to a dataset in a microservice. This sample includes two infotables, the datasetRef infotable (which, in turn, contains the AnalyticsDatasetMetadata infotable) and the learners infotable. For more information about the structure of the training request and the infotables, see the
ThingWorx Analytics API documentation.
{
"jobName": "beanPro Grinder Error model",
"validationHoldoutPercentage": 0.2,
"maxAllowedFields": 50,
"goalField": "Grinder Error Occurrence",
"description": "predicting the grinder error occurence value on the beanpro
dataset",
"datasetRef": {
"created": 1501687620663,
"description": "",
"name": "Infotable",
"dataShape": {
"fieldDefinitions": {
"filter": {
"name": "filter",
"aspects": {},
"description": "Refine the dataset by filtering out values",
"baseType": "STRING",
"ordinal": 3
},
"datasetUri": {
"name": "datasetUri",
"aspects": {},
"description": "Location of the data",
"baseType": "STRING",
"ordinal": 1
},
"metadata": {
"name": "metadata",
"aspects": {
"dataShape": "AnalyticsDatasetMetadata"
},
"description": "Metadata describing the locally provided data",
"baseType": "INFOTABLE",
"ordinal": 6
},
"data": {
"name": "data",
"aspects": {},
"description": "Provide the data locally as part of the request",
"baseType": "STRING",
"ordinal": 5
},
"format": {
"name": "format",
"aspects": {},
"description": "Structure of the data (eg: csv, json)",
"baseType": "STRING",
"ordinal": 2
},
"exclusions": {
"name": "exclusions",
"aspects": {
"dataShape": "GenericStringList"
},
"description": "Refine the dataset by removing fields",
"baseType": "INFOTABLE",
"ordinal": 4
}
},
"name": "AnalyticsDatasetRef",
"description": "Pointer to an analytics dataset"
},
"rows": [
{
"datasetUri": "dataset:///c2b17f4b-6273-4e04-a466-26f0755e8af3",
"metadata": {
"created": 1501687665676,
"description": "",
"name": "Infotable",
"dataShape": {
"fieldDefinitions": {
"timeSamplingInterval": {
"name": "timeSamplingInterval",
"aspects": {},
"description": "Time in between observations in Temporal field",
"baseType": "INTEGER",
"ordinal": 7
},
"isStatic": {
"name": "isStatic",
"aspects": {},
"description": "If this field remains unchanged over time and
shouldn't be pivoted for time series transformations",
"baseType": "BOOLEAN",
"ordinal": 8
},
"fieldName": {
"name": "fieldName",
"aspects": {},
"description": "Name of the field",
"baseType": "STRING",
"ordinal": 1
},
"min": {
"name": "min",
"aspects": {},
"description": "Minimum observed value for Continuous fields",
"baseType": "NUMBER",
"ordinal": 4
},
"max": {
"name": "max",
"aspects": {},
"description": "Maximum observed value for Continuous fields",
"baseType": "NUMBER",
"ordinal": 5
},
"dataType": {
"name": "dataType",
"aspects": {},
"description": "Format of the data (STRING,DOUBLE,BOOLEAN,INTEGER)",
"baseType": "STRING",
"ordinal": 2
},
"opType": {
"name": "opType",
"aspects": {},
"description": "Behavior of the data (CONTINUOUS,CATEGORICAL,ORDINAL,
BOOLEAN,TEMPORAL,ENTITY_ID)",
"baseType": "STRING",
"ordinal": 3
},
"values": {
"name": "values",
"aspects": {
"dataShape": "GenericStringList"
},
"description": "Collection of possible values for Ordinal and
Categorical fields",
"baseType": "INFOTABLE",
"ordinal": 6
}
},
"name": "AnalyticsDatasetMetadata",
"description": "Metadata describing a dataset field"
},
"rows": []
},
"format": "parquet",
"exclusions": {
"created": 1501687665679,
"description": "",
"name": "Infotable",
"dataShape": {
"fieldDefinitions": {
"item": {
"name": "item",
"aspects": {
"isPrimaryKey": true
},
"description": "Item",
"baseType": "STRING",
"ordinal": 0
}
},
"name": "GenericStringList",
"description": "Generic data shape to hold a list of strings"
},
"rows": []
}
}
]
},
"learners": {
"created": 1501687620656,
"description": "",
"name": "Infotable",
"dataShape": {
"fieldDefinitions": {
"maxDepth": {
"name": "maxDepth",
"aspects": {},
"description": "The maximum tree depth (DECISION_TREE, RANDOM_FOREST,
and GRADIENT_BOOST only)",
"baseType": "INTEGER",
"ordinal": 2
},
"learningTechnique": {
"name": "learningTechnique",
"aspects": {},
"description": "The technique to use for learning (NEURAL_NET,
LINEAR_REGRESSION, LOGISTIC_REGRESSION, DECISION_TREE,
RANDOM_FOREST, and GRADIENT_BOOST)",
"baseType": "STRING",
"ordinal": 1
},
"hiddenUnitPercentage": {
"name": "hiddenUnitPercentage",
"aspects": {},
"description": "The percentage of the number of input nodes to use
in each hidden layer of a NEURAL_NET",
"baseType": "NUMBER",
"ordinal": 5
},
"layerCount": {
"name": "layerCount",
"aspects": {},
"description": "The number of layers to use in NEURAL_NET",
"baseType": "INTEGER",
"ordinal": 4
},
"numberOfIterations": {
"name": "numberOfIterations",
"aspects": {},
"description": "The number of iterations for GRADIENT_BOOST",
"baseType": "INTEGER",
"ordinal": 6
},
"treeCount": {
"name": "treeCount",
"aspects": {},
"description": "The number of trees to use in RANDOM_FOREST",
"baseType": "INTEGER",
"ordinal": 3
},
"removeDuplicatesAndUniformColumns": {
"name": "removeDuplicatesAndUniformColumns",
"aspects": {},
"description": "Determines whether to remove duplicate columns for
DECISION_TREE",
"baseType": "BOOLEAN",
"ordinal": 6
}
},
"name": "AnalyticsTrainingLearner",
"description": "The configuration of a machine learning model for a
training job"
},
"rows": [
{
"learningTechnique": "NEURAL_NET"
}
]
}
}