# You can also use wget
curl-XDELETEhttps://app.datarobot.com/api/v2/dataSlices/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/dataSlices/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/dataSlices/{dataSliceId}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/dataSlices/{dataSliceId}/sliceSizes/?projectId=string&source=backtest_0\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
query
string
true
The project ID.
source
query
string
true
The source of data to use to calculate the size.
externalDatasetId
query
string
false
The external dataset ID to use when calculating the size of a slice. Use this parameter only when the source is 'externalTestSet'.
modelId
query
string
false
The model ID whose training dataset should be sliced. Use this parameter only when the source is 'training'.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/dataSlices/{dataSliceId}/sliceSizes/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/insights/featureEffects/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/insights/featureEffects/models/{entityId}/?limit=10&offset=0&source=validation\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
limit
query
integer
true
The numbers of items to return.
offset
query
integer
true
The number of items to skip before starting to collect the result set.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/insights/featureImpact/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/insights/featureImpact/models/{entityId}/?limit=10&offset=0\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
limit
query
integer
true
The numbers of items to return.
offset
query
integer
true
The number of items to skip before starting to collect the result set.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/insights/liftChart/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/insights/liftChart/models/{entityId}/?limit=10&offset=0\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
limit
query
integer
true
The numbers of items to return.
offset
query
integer
true
The number of items to skip before starting to collect the result set.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/insights/residuals/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/insights/residuals/models/{entityId}/?limit=10&offset=0\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
limit
query
integer
true
The numbers of items to return.
offset
query
integer
true
The number of items to skip before starting to collect the result set.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/insights/rocCurve/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/insights/rocCurve/models/{entityId}/?limit=10&offset=0\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
limit
query
integer
true
The numbers of items to return.
offset
query
integer
true
The number of items to skip before starting to collect the result set.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/multilabelInsights/{multilabelInsightsKey}/histogram/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
multilabelInsightsKey
path
string
true
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/multilabelInsights/{multilabelInsightsKey}/pairwiseManualSelections/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
multilabelInsightsKey
path
string
true
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/multilabelInsights/{multilabelInsightsKey}/pairwiseManualSelections/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
# You can also use wget
curl-XDELETEhttps://app.datarobot.com/api/v2/multilabelInsights/{multilabelInsightsKey}/pairwiseManualSelections/{manualSelectionListId}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
multilabelInsightsKey
path
string
true
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
# You can also use wget
curl-XPATCHhttps://app.datarobot.com/api/v2/multilabelInsights/{multilabelInsightsKey}/pairwiseManualSelections/{manualSelectionListId}/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/multilabelInsights/{multilabelInsightsKey}/pairwiseStatistics/¶
Retrieve multilabel specific pairwise label statistics for the given multilabel insights key: correlation, joint probability and conditional probability.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/multilabelInsights/{multilabelInsightsKey}/pairwiseStatistics/?statisticType=conditionalProbability\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
statisticType
query
string
true
Type of pairwise statistic.
multilabelInsightsKey
path
string
true
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/anomalyAssessmentRecords/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
modelId
query
string
false
The model ID to filter records by.
backtest
query
any
false
The backtest to filter records by.
source
query
string
false
The source of the data to filter records by.
seriesId
query
string
false
Can be specified for multiseries projects. The series id to filter records by.
# You can also use wget
curl-XDELETEhttps://app.datarobot.com/api/v2/projects/{projectId}/anomalyAssessmentRecords/{recordId}/\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/anomalyAssessmentRecords/{recordId}/explanations/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
startDate
query
string(date-time)
false
The start of the date range to return. Date should be in UTC format. For example: 2019-08-01T00:00:00.000000Z.
endDate
query
string(date-time)
false
The end of the date range to return, inclusive. Date should be in UTC format. For example: 2020-01-01T00:00:00.000000Z.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/anomalyAssessmentRecords/{recordId}/predictionsPreview/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/biasVsAccuracyInsights/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
accuracyMetric
query
string
false
The metric to return model accuracy scores. Defaults to the optimization metric configured in project options.
protectedFeature
query
any
false
Name of the protected feature.
fairnessMetric
query
any
false
The fairness metric used to calculate the fairness scores.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/dataSlices/?limit=100&offset=0\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
limit
query
integer
true
The numbers of items to return.
offset
query
integer
true
The number of items to skip before starting to collect the result set.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/accuracyOverTimePlots/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided an average plot for the first 1000 series will be retrieved.
backtest
query
any
false
Retrieve plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
source
query
string
false
The source of the data for the backtest/holdout.
forecastDistance
query
integer
false
Forecast distance to retrieve the data for. If not specified, the first forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
resolution
query
string
false
Specifying at which resolution the data should be binned. If not specified, optimal resolution will be used to build chart data with number of bins <= maxBinSize
maxBinSize
query
integer
false
Specifies the maximum number of bins for the retrieval.
startDate
query
string(date-time)
false
The start of the date range to return. If not specified, start date for requested plots will be used.
endDate
query
string(date-time)
false
The end of the date range to return. If not specified, end date for requested plots will be used.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/accuracyOverTimePlots/metadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
forecastDistance
query
integer
false
Forecast distance to retrieve the data for. If not specified, the first forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided a metadata of average plot for the first 1000 series will be retrieved.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/accuracyOverTimePlots/preview/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided an average plot for the first 1000 series will be retrieved.
backtest
query
any
false
Retrieve plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
source
query
string
false
The source of the data for the backtest/holdout.
forecastDistance
query
integer
false
Forecast distance to retrieve the data for. If not specified, the first forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/anomalyOverTimePlots/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided an average plot for the first 1000 series will be retrieved.
backtest
query
any
false
Retrieve plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
source
query
string
false
The source of the data for the backtest/holdout.
resolution
query
string
false
Specifying at which resolution the data should be binned. If not specified, optimal resolution will be used to build chart data with number of bins <= maxBinSize
maxBinSize
query
integer
false
Specifies the maximum number of bins for the retrieval.
startDate
query
string(date-time)
false
The start of the date range to return. If not specified, start date for requested plots will be used.
endDate
query
string(date-time)
false
The end of the date range to return. If not specified, end date for requested plots will be used.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/anomalyOverTimePlots/metadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided a metadata of average plot for the first 1000 series will be retrieved.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/anomalyOverTimePlots/preview/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided an average plot for the first 1000 series will be retrieved.
backtest
query
any
false
Retrieve plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
source
query
string
false
The source of the data for the backtest/holdout.
predictionThreshold
query
number
false
Only bins with predictions exceeding this threshold will be returned in the response.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/backtestStabilityPlot/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
metricName
query
string
false
The name of the metric to retrieve the scores for. If omitted, the default project metric will be used
forecastDistance
query
integer
false
The forecast distance to retrieve the plot for. If not specified, the scores for each partition are aggregated across all forecast distances. This parameter is only available for time series models.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/datetimeModels/{modelId}/datasetAccuracyOverTimePlots/{datasetId}/¶
Retrieve the Accuracy Over Time (AOT) chart data for an external dataset for a project.
Datetimes are specified in accordance with :ref:timestamp and duration formatting rules <time_format>.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/datasetAccuracyOverTimePlots/{datasetId}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
maxBinSize
query
integer
false
The limit of returned bins.
startDate
query
string(date-time)
false
The start of the date range to return (UTC string), for example: '2010-05-13T00:00:00.000000Z'. If not specified, the start date for this model and source of the data will be used instead.
endDate
query
string(date-time)
false
The end of the date range to return (UTC string), for example: '2010-05-13T00:00:00.000000Z'. If not specified, the end date for this model and source of the data will be used instead.
resolution
query
string
false
Specifies at which resolution the data should be binned. If not specified, optimal resolution will be used to build chart data such that bins <= maxBinSize.
projectId
path
string
true
The project id which was used to compute the AOT chart.
modelId
path
string
true
The model id which was used to compute the AOT chart.
datasetId
path
string
true
The dataset id which was used to compute the AOT chart.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/datetimeModels/{modelId}/datasetAccuracyOverTimePlots/{datasetId}/metadata/¶
Retrieve the metadata of the Accuracy Over Time (AOT) chart for an external dataset.
Datetimes are specified in accordance with :ref:timestamp and duration formatting rules <time_format>.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/datasetAccuracyOverTimePlots/{datasetId}/metadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
path
string
true
The project id which was used to compute the AOT chart.
modelId
path
string
true
The model id which was used to compute the AOT chart.
datasetId
path
string
true
The dataset id which was used to compute the AOT chart.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/datetimeModels/{modelId}/datasetAccuracyOverTimePlots/{datasetId}/preview/¶
Retrieve a preview of the Accuracy Over Time (AOT) chart for an external dataset.
Datetimes are specified in accordance with :ref:timestamp and duration formatting rules <time_format>.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/datasetAccuracyOverTimePlots/{datasetId}/preview/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
path
string
true
The project id which was used to compute the AOT chart.
modelId
path
string
true
The model id which was used to compute the AOT chart.
datasetId
path
string
true
The dataset id which was used to compute the AOT chart.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/datetimeModels/{modelId}/datetimeTrendPlots/¶
Computes Datetime Trend plots for time series and OTV projects:
* For OTV projects, computes Accuracy over Time plots.
* For time series supervised projects, computes both Accuracy over Time plots and Forecast vs Actual plots.
.. minversion:: v2.25
* For unsupervised time series and OTV models, computes Anomaly Over Time plots.
.. note::
For the multiseries time series projects only first 1000 series in alphabetical order and an average plot for them will be computed.
.. note::
Maximum 100 forecast distances can be requested for calculation in time series supervised projects.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/datetimeTrendPlots/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/datetimeModels/{modelId}/featureEffects/¶
Retrieve Feature Effects for a model backtest.
Feature Effects provides partial dependence and predicted vs actual values for the top 500 features, ordered by feature impact score.
The partial dependence shows marginal effect of a feature on the target variable after accounting for the average effects of all other predictive features. It indicates how, holding all other variables except the feature of interest as they were, the value of this feature affects your prediction.
If a Feature Effects job was previously submitted for a given backtest, this endpoint will return a response structured as {"message": , "jobId": } where jobId is the ID of the job. Retrieve the job with GET /api/v2/projects/{projectId}/jobs/{jobId}/
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/featureEffects/?backtestIndex=string\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
source
query
string
false
Models data source.
backtestIndex
query
string
true
The backtest index. For example: 0, 1, ..., 20, holdout, startstop.
Project, model, source, backtest index, or computation results do not exist.
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/datetimeModels/{modelId}/featureEffects/¶
Add a request to the queue to calculate Feature Effects for a backtest.
If the job has been previously submitted, the request fails, returning the jobId of the previously submitted job. Use this jobId to check status of the previously submitted job.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/featureEffects/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/datetimeModels/{modelId}/featureEffectsMetadata/¶
Retrieve Feature Effects metadata for each backtest. Response contains status and available sources for each backtest of the model.
One of the provided backtestIndex indexes used for submitting the compute request and retrieving Feature Effects.
* Start/stop models contain a single backtestIndex response value of startstop.
* Other models contain backtestIndex of 0, 1, ..., holdout.
One of the provided source parameters used for retrieving Feature Effects.
* Each backtest source can be, at a minimum, training or validation. If holdout is configured for the project, backtestIndex also includes holdout with sources training and holdout.
* Source value of training is always available. (versions prior to v2.17 support validation only)
* When a start/stop model is trained into validation or holdout without stacked predictions (i.e., no out-of-sample predictions in validation or holdout), validation and holdout sources are not available.
* Source holdout is not available when there is no holdout configured for the project.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/featureEffectsMetadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/datetimeModels/{modelId}/forecastDistanceStabilityPlot/¶
Retrieve a plot displaying the stability of the time series model across different forecast distances.
.. note::
All durations and datetimes are specified in accordance with the :ref:timestamp and duration formatting rules <time_format>.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/forecastDistanceStabilityPlot/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
metricName
query
string
false
The name of the metric to retrieve the scores for. If omitted, the default project metric will be used.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/forecastVsActualPlots/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided an average plot for the first 1000 series will be retrieved.
backtest
query
any
false
Retrieve plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
source
query
string
false
The source of the data for the backtest/holdout.
resolution
query
string
false
Specifying at which resolution the data should be binned. If not specified, optimal resolution will be used to build chart data with number of bins <= maxBinSize
forecastDistanceStart
query
integer
false
The start of forecast distance range (forecast window) to retrieve. If not specified, the first forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
forecastDistanceEnd
query
integer
false
The end of forecast distance range (forecast window) to retrieve. If not specified, the last forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
maxBinSize
query
integer
false
Specifies the maximum number of bins for the retrieval.
startDate
query
string(date-time)
false
The start of the date range to return. If not specified, start date for requested plots will be used.
endDate
query
string(date-time)
false
The end of the date range to return. If not specified, end date for requested plots will be used.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/forecastVsActualPlots/metadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided a metadata of average plot for the first 1000 series will be retrieved.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/forecastVsActualPlots/preview/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
seriesId
query
string
false
The name of the series to retrieve. Only available for time series multiseries projects. If not provided an average plot for the first 1000 series will be retrieved.
backtest
query
any
false
Retrieve plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/multiclassFeatureEffects/?backtestIndex=string\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
source
query
string
false
Models data source.
backtestIndex
query
string
true
The backtest index. For example: 0, 1, ..., 20, holdout, startstop.
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
Project, model, source or computation results do not exist.
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/datetimeModels/{modelId}/multiclassFeatureEffects/¶
Compute feature effects for a multiclass datetime model. If the job has been previously submitted, the request fails, returning the jobId of the previously submitted job. Use this jobId to check status of the previously submitted job.
NOTE: feature effects are computed for top 100 classes.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/multiclassFeatureEffects/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
Queue submission error. If the rowCount exceeds the maximum or minimum value for this dataset. Minimum is 10 rows. Maximum is 100000 rows or the training sample size of the model, whichever is less. If invalid class names are provided in classes.If neither features nor topNFeatures is provided. If invalid backtestIndex is provided.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/multiseriesHistograms/?attribute=rowCount&bins=10\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
attribute
query
string
true
The series attribute to build a histogram for.
metric
query
string
false
The name of the metric to retrieve the histogram for attributes "validationScore", "backtestingScore", and"holdoutScore". If omitted, the default project metric will be used.
bins
query
string
true
The number of bins in a histogram. Can be 10, 20 or 50. The default is 10.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/multiseriesScores/?offset=0&limit=100\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
multiseriesValue
query
string
false
Only the series containing the given value in one of the series ID columns will be returned if specified.
offset
query
integer
true
The number of results to skip. Defaults to 0 if not specified.
limit
query
integer
true
The maximum number of results to return. Defaults to 100 if not specified.
metric
query
string
false
The name of the metric to retrieve the scores for.If omitted, the default project metric will be used.
orderBy
query
string
false
Used for sorting the series. Supported attributes for ordering include: "multiseriesValue", "rowCount", "validationScore", "holdoutScore" and "backtestingScore", "startDate", "endDate", and "targetAverage".Prefix the attribute name with a dash to sort in descending order,e.g. orderBy=-rowCount. If multiple series with equal values of the ordering attributeexist, ties will be broken arbitrarily.
filterBy
query
string
false
Used to specify on which attribute values to filter the series.Supported attributes for filtering include: "rowCount", "startDate", "endDate", "targetAverage", "validationScore", "holdoutScore", and "backtestingScore".filterByBins and numberOfBins are required if this parameter is used.
numberOfBins
query
string
false
Used to specify the number of bins in the histogram on which to filter the series.Can be 10, 20 or 50.filterBy and filterByBins are required if this parameter is used.
filterByBins
query
string
false
Used to specify the multiseries histogram bins on which to filter the series.filterBy and numberOfBins are required if this parameter is used.
clusterNames
query
string
false
Used to specify the specific cluster on which to filter the series.filterBy is required if this parameter is used.Only valid for unsupervised clustering projects.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/datetimeModels/{modelId}/multiseriesScores/¶
Request the computation of per-series scores for a multiseries model.
.. note::
Computation uses available partitions only. This endpoint will not compute backtesting scores if no backtesting scores exist prior to this request.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/multiseriesScores/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/datetimeModels/{modelId}/multiseriesScores/file/\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
multiseriesValue
query
string
false
If specified, only the series containing the given value in one of the series ID columns will be returned.
metric
query
string
false
The name of the metric to retrieve the scores for. If omitted, the default project metric will be used.
List of scores on prediction datasets for a project with filtering option by dataset or model or both of them. Prediction dataset may have scores if it contained a column with actual values and predictions were computed on this dataset.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/externalScores/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
datasetId
query
string
false
If provided will return scores for dataset with matching datasetId.
modelId
query
string
false
If provided will return scores for model with matching modelId.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/externalScores/¶
Compute model scores for external dataset, first upload your dataset to the project, and then using the corresponding datasetId, compute scores against that dataset. Computing external scores and insights depends on computed prediction, predictions will be computed if they are not available for this dataset. In order to compute scores and insights, uploaded dataset should contain actual value column. This api is not available in time series projects.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/externalScores/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/featureAssociationFeaturelists/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/featureAssociationMatrix/?type=association&metric=mutualInfo\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
type
query
string
true
the type of dependence for the data. Must be either association or correlation. Since v2.19 this is optional and defaults to association.
metric
query
string
true
the name of a metric to get pairwise data for. Must be one of mutualInfo, cramersV, spearman, pearson, or tau. Since v2.19 this is optional and defaults to mutualInfo.
featurelistId
query
string
false
the feature list to lookup FAM data for. By default, depending on the type of the project Informative Features or Timeseries Informative Features list will be used.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/featureAssociationMatrix/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/featureAssociationMatrixDetails/?feature1=string&feature2=string\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
feature1
query
string
true
The name of a feature.
feature2
query
string
true
the name of another feature.
featurelistId
query
string
false
the feature list to lookup FAM data for. By default, depending on the type of the project Informative Features or Timeseries Informative Features list will be used.
This project does not support feature associations, (e.g. multilabel, multiseries, time series unsupervised projects.).
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/features/{featureName}/frequentValues/¶
Retrieve the frequent values information for a particular feature.
Only valid for numeric features.
This route returns information about the frequent values seen for a particular feature,
based on the EDA sample of the dataset. Up to 60 values will be returned,
and when more values are present, they will be bucketed into a level called "==All Other=="
at the end of the response.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/features/{featureName}/frequentValues/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/geometryFeaturePlots/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/geometryFeaturePlots/{featureName}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
path
string
true
Project Id. It is the project to select the feature from.
featureName
path
string
true
Name of location feature to plot on map. Must be supplied in order to determine which plot to retrieve.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/models/{modelId}/anomalyAssessmentInitialization/¶
Initialize the anomaly assessment insight and calculate Shapley explanations for the most anomalous points in the subset. The insight is available for anomaly detection models in time series unsupervised projects which also support calculation of Shapley values.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/anomalyAssessmentInitialization/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/anomalyInsightsFile/¶
Retrieve a CSV file of the raw data displayed with the anomaly score from the specific model. The number of rows included will be set by the expected outlier fraction but up to a maximum of 1000 rows. Only models built from anomaly detection blueprints have those insights.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/anomalyInsightsFile/\-H"Authorization: Bearer {access-token}"
project Id / model Id does not exist or model doesn't have anomaly insights table.
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/anomalyInsightsTable/¶
Retrieve a table of the raw data displayed with the anomaly score from the specific model. The number of rows displayed is limited to 100 rows by the ANOMALY_INSIGHT_SAMPLE_ROW_COUNT configuration setting. Additionally, feature column count and the size of data in text fields is also limited. Only models built from anomaly detection blueprints have those insights.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/anomalyInsightsTable/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
anomalyScoreRounding
query
integer
false
number of decimals each element anomalyScore column will be rounded to.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/clusterInsights/?limit=15\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
true
At most this many results are returned. The default may change without notice.
orderBy
query
string
false
Order results by the specified field value.
searchFor
query
string
false
Search for a specific string in a feature name.This search is case insensitive. If not specified, all features will be returned.
Feature Impact is required. Please, compute it first.
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/models/{modelId}/clusterInsights/¶
Compute Cluster Insights for a clustering project model.The number of features computed for cluster insights are capped at 100, starting with the features used to train the model sorted by feature impact (high to low), and then the remaining features in the dataset alphabetically.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/clusterInsights/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/clusterInsights/download/\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
format
query
string
false
A format to use.
featurelistId
query
string
false
The ID of the featurelist to download data for. If not specified all columns will be downloaded.
Contains an auto generated filename for this download ("attachment;filename=cluster_insights__.csv").
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/confusionCharts/¶
Retrieve all available confusion charts for model. The response will
include a json array of all available confusion charts, in the same format as the response
from [GET /api/v2/projects/{projectId}/models/{modelId}/confusionCharts/{source}/][get-apiv2projectsprojectidmodelsmodelidconfusionchartssource].
.. note:: Available for multiclass projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/confusionCharts/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/confusionCharts/{source}/¶
Retrieve the confusion chart data from a single source. A confusion chart consists of the confusion matrix for all classes, classes frequencies and `oneVsAll` metrics for all classes. The confusion matrix can be requested in a particular sort order and orientated by rows or columns. A subset of the confusion matrix can also be requested in part by specifying slicing indices. Throughout the following specification, `C` refers to the total number of classes in the dataset. The full confusion matrix refers to the confusion matrix with `C` classes.
.. note:: Available for multiclass projects only.
An example on the meaning of wasActualPercentages and wasPredictedPercentages:
Let's say we have the following data:
.. code-block:: js
classMetrics.classA.wasActualPercentages[0].percentage = 0.56
classMetrics.classA.wasPredictedPercentages[0].percentage = 0.62
classA.wasActualPercentages[0].otherClassName = "classB"
classA.wasPredictedPercentages[0].otherClassName = "classB"
That means:
1) "Given that it was actually classA, it predicted classB 56% of the time".
2) "Given that classA was predicted, it was actually classB 62% of the time".
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/confusionCharts/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
orderBy
query
string
false
Ordering the chart data by following attributes.Prefix the attribute name with a dash to sort in descending order, e.g. orderBy='-predictedCount'
orientation
query
string
false
Determines whether the values in the rows of the confusion matrix should correspond to the same actual class ('actual') or predicted class ('predicted').
rowStart
query
integer
false
start index of row for slicing the confusion matrix.
rowEnd
query
integer
false
end index of row for slicing the confusion matrix.
colStart
query
integer
false
start index of column for slicing the confusion matrix.
colEnd
query
integer
false
end index of column for slicing the confusion matrix.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/confusionCharts/{source}/classDetails/¶
Calculates and sends frequency of class in distributed among other
classes for actual and predicted data. A confusion chart class details for given class gives
stats of misclassification done by model for given class for actual and predicted data.
.. note:: Available for multiclass projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/confusionCharts/{source}/classDetails/?className=string\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
className
query
string
true
Name of a class for which distribution frequency is requested.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/confusionCharts/{source}/metadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
orderBy
query
string
false
Ordering the chart data by following attributes.Prefix the attribute name with a dash to sort in descending order, e.g. orderBy='-predictedCount'
orientation
query
string
false
Determines whether the values in the rows of the confusion matrix should correspond to the same actual class ('actual') or predicted class ('predicted').
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/crossClassAccuracyScores/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of items to skip. Defaults to 0 if not provided.
limit
query
integer
false
Number of items to return, defaults to 100 if not provided.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/crossClassAccuracyScores/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/dataDisparityInsights/?feature=string&className1=string&className2=string\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of items to skip. Defaults to 0 if not provided.
limit
query
integer
false
Number of items to return, defaults to 100 if not provided.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/dataDisparityInsights/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/¶
List of Confusion Charts objects on external datasets for a project with filtering option by dataset. Prediction dataset may have Confusion Chart for multiclass projects computed if it contained a target with actual values and insights were computed on this dataset. A confusion chart consists of the confusion matrix for all classes, classes frequencies and oneVsAll metrics for all classes. The confusion matrix can be requested in a particular sort order and orientated by rows or columns. Available for multiclass projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
orderBy
query
string
false
Ordering the chart data by following attributes.Prefix the attribute name with a dash to sort in descending order, e.g. orderBy='-predictedCount'
orientation
query
string
false
Determines whether the values in the rows of the confusion matrix should correspond to the same actual class ('actual') or predicted class ('predicted').
rowStart
query
integer
false
start index of row for slicing the confusion matrix.
rowEnd
query
integer
false
end index of row for slicing the confusion matrix.
colStart
query
integer
false
start index of column for slicing the confusion matrix.
colEnd
query
integer
false
end index of column for slicing the confusion matrix.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/{datasetId}/¶
Retrieve Confusion Chart objects on external datasets for a project. Prediction dataset may have Confusion Chart for multiclass projects computed if it contained a target with actual values and insights were computed on this dataset. A confusion chart consists of the confusion matrix for all classes, classes frequencies and oneVsAll metrics for all classes. The confusion matrix can be requested in a particular sort order and oriented by rows or columns (zero-indexed). Available for multiclass projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/{datasetId}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
orderBy
query
string
false
Ordering the chart data by following attributes.Prefix the attribute name with a dash to sort in descending order, e.g. orderBy='-predictedCount'
orientation
query
string
false
Determines whether the values in the rows of the confusion matrix should correspond to the same actual class ('actual') or predicted class ('predicted').
rowStart
query
integer
false
start index of row for slicing the confusion matrix.
rowEnd
query
integer
false
end index of row for slicing the confusion matrix.
colStart
query
integer
false
start index of column for slicing the confusion matrix.
colEnd
query
integer
false
end index of column for slicing the confusion matrix.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/{datasetId}/classDetails/¶
Calculate and sends frequency of class in distributed among other classes for actual and predicted data. A confusion chart class details for given class gives stats of misclassification done by model for given class for actual and predicted data. Available for multiclass projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/{datasetId}/classDetails/?className=string\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
className
query
string
true
Name of a class for which distribution frequency is requested.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetConfusionCharts/{datasetId}/metadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
orderBy
query
string
false
Ordering the chart data by following attributes.Prefix the attribute name with a dash to sort in descending order, e.g. orderBy='-predictedCount'
orientation
query
string
false
Determines whether the values in the rows of the confusion matrix should correspond to the same actual class ('actual') or predicted class ('predicted').
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetLiftCharts/¶
List of Lift chart objects on prediction datasets for a project with filtering option by dataset. Prediction dataset may have Lift chart computed if it contained a column with actual values and predictions were computed on this dataset. This controller is not supported for multiclass classification projects. For multiclass, instead use /projects//models//datasetMulticlassLiftCharts/.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetLiftCharts/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
datasetId
query
string
false
If provided will return Lift chart for dataset with matching datasetId.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetMulticlassLiftCharts/¶
List of Multiclass Lift chart objects on prediction datasets for a project with filtering option by dataset. Prediction dataset may have Multiclass Lift chart computed if it contained a column with actual values and predictions were computed on this dataset. Multiclass Lift charts are supported for multiclass classification projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetMulticlassLiftCharts/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
datasetId
query
string
false
If provided will return Lift chart for dataset with matching datasetId.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetResidualsCharts/¶
List of residuals charts objects on prediction datasets for a project with filtering option by dataset. Prediction dataset may have residuals chart computed if it contained a column with actual values and predictions were computed on this dataset.
Residuals charts are supported for regression projects only.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetResidualsCharts/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
datasetId
query
string
false
If provided will return ROC curve for dataset with matching datasetId.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/datasetRocCurves/¶
List of ROC curve objects on prediction datasets for a project with filtering option by dataset.
Prediction dataset may have ROC curve computed if it contained a column with actual values and predictions were computed on this dataset.
Each ROC curve object includes an array of points showing the performance of the model at different thresholds for classification, and arrays of sample predictions for both the positive and negative classes.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/datasetRocCurves/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
datasetId
query
string
false
If provided will return ROC curve for dataset with matching datasetId.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/fairnessInsights/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of items to skip. Defaults to 0 if not provided.
limit
query
integer
false
Number of items to return, defaults to 100 if not provided.
fairnessMetricsSet
query
string
false
Metric to use for calculating fairness. Can be one of proportionalParity, equalParity, predictionBalance, trueFavorableAndUnfavorableRateParity or FavorableAndUnfavorablePredictiveValueParity. Used and required only if Bias & Fairness in AutoML feature is enabled.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/fairnessInsights/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/featureEffects/¶
Retrieve Feature Effects for the model.
Feature Effects provides partial dependence and predicted vs actual values for the top 500 features, ordered by feature impact score.
The partial dependence shows marginal effect of a feature on the target variable after accounting for the average effects of all other predictive features. It indicates how, holding all other variables except the feature of interest as they were, the value of this feature affects your prediction.
If a Feature Effects job was previously submitted, this endpoint will return a response structured as {"message": , "jobId": } where jobId is the ID of the job. Retrieve the job with GET /api/v2/projects/{projectId}/jobs/{jobId}/
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/featureEffects/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Project, model, source or computation results do not exist.
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/models/{modelId}/featureEffects/¶
Add a request to the queue to calculate Feature Effects.
If the job has been previously submitted, the request fails, returning the jobId of the previously submitted job. Use this jobId to check status of the previously submitted job.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/featureEffects/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
If the rowCount exceeds the maximum or minimum value for this dataset. Minimum
is 10 rows. Maximum is 100000 rows or the training sample size of the model,
whichever is less.|None|
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/featureEffectsMetadata/¶
Retrieve Feature Effects metadata. Response contains status and available sources.
One of the provided source parameters used for retrieving Feature Effects.
* Source can be, at a minimum, training or validation. If holdout is configured for the project, source also includes holdout.
* Source value of training is always available. (versions prior to v2.17 support validation only)
* When a model is trained into validation or holdout without stacked predictions (i.e., no out-of-sample predictions in validation or holdout), validation and holdout sources are not available.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/featureEffectsMetadata/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/featureImpact/¶
Retrieve feature impact scores for features in a model.
Feature Impact is computed for each column by creating new data with that column randomly permuted (but the others left unchanged), and seeing how the error metric score for the predictions is affected. Elsewhere this technique is sometimes called 'Permutation Importance'.
The impactUnnormalized is how much worse the error metric score is when making predictions on this modified data. The impactNormalized is normalized so that the largest value is 1. In both cases, larger values indicate more important features. If a feature is a redundant feature, i.e. once other features are considered it doesn't contribute much in addition, the redundantWith value is the name of feature that has the highest correlation with this feature.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/featureImpact/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
backtest
query
any
false
The backtest value used for Feature Impact computation. Applicable for datetime aware models.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/models/{modelId}/featureImpact/¶
Add a request to calculate feature impact to the queue.
If the job has been previously submitted, the request will fail and return the jobId of previously submitted job. This jobId can be used to check status of previously submitted job.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/featureImpact/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
If feature impact has already run will return error including jobId property which is the jobId of the previously started feature impact job.
.. minversion:: v2.21
If the rowCount exceeds the maximum or minimum value for this dataset. Minimum is 10 rows. Maximum is 100000 rows or the training sample size of the model, whichever is less.|None|
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/featureLists/{datasetId}/clusterInsights/?limit=15\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
true
At most this many results are returned. The default may change without notice.
orderBy
query
string
false
Order results by the specified field value.
searchFor
query
string
false
Search for a specific string in a feature name.This search is case insensitive. If not specified, all features will be returned.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/labelwiseRocCurves/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
labels
query
string
false
Labels for which data is requested.
searchQuery
query
string
false
Search query for label.
sortBy
query
string
false
Property to sort labels in the response.
sortOrder
query
string
false
Sort order.
threshold
query
number
false
Threshold at which the metric should be sorted.
offset
query
integer
false
Number of labels to skip.
limit
query
integer
false
Number of labels to return.
includeModelAverage
query
boolean
false
Whether model average metrics should be included in the response.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/liftChart/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/liftChart/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/multiclassFeatureEffects/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
source
query
string
false
Models data source.
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
Project, model, source or computation results do not exist.
None
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/models/{modelId}/multiclassFeatureEffects/¶
Compute feature effects for a multiclass model. If the job has been previously submitted, the request fails, returning the jobId of the previously submitted job. Use this jobId to check status of the previously submitted job.
NOTE: feature effects are computed for top 100 classes.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/multiclassFeatureEffects/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
Queue submission error. If the rowCount exceeds the maximum or minimum value for this dataset. Minimum is 10 rows. Maximum is 100000 rows or the training sample size of the model, whichever is less. If neither features nor topNFeatures is provided.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/multiclassFeatureImpact/¶
Retrieve feature impact scores for each class in a multiclass model.
Feature Impact is computed for each column by creating new data with that column randomly permuted (but the others left unchanged), and seeing how the error metric score for the predictions is affected. Elsewhere this technique is sometimes called 'Permutation Importance'.
The impactUnnormalized is how much worse the error metric score is when making predictions on this modified data. The impactNormalized is normalized so that the largest value is 1. In both cases, larger values indicate more important features.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/multiclassFeatureImpact/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/multiclassLiftChart/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/multiclassLiftChart/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/multilabelLiftCharts/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XDELETEhttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/predictionExplanationsInitialization/\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/predictionExplanationsInitialization/¶
Retrieve the current PredictionExplanationsInitialization.
A PredictionExplanationsInitialization is a pre-requisite for successfully computing prediction explanations using a particular model, and can be used to preview the prediction explanations that would be generated for a complete dataset.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/predictionExplanationsInitialization/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/predictionExplanationsInitialization/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/residuals/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/residuals/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/rocCurve/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/rocCurve/{source}/¶
Retrieve the ROC curve data from a single source. The response includes an array of pointsshowing the performance of the model at different thresholds for classification, and arrays of sample predictions for both the positive and negative classes.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/rocCurve/{source}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/models/{modelId}/shapImpact/¶
Retrieve Feature Impact for a model. Shap impact is computed by calculating the shap values on a sample of training data and then taking the mean absolute value for each column. The larger value of impact indicate more important feature.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/shapImpact/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/shapImpact/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/models/{modelId}/wordCloud/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
excludeStopWords
query
string
false
Set to true if you want stopwords excluded from the response.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/multicategoricalInvalidFormat/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
path
string
true
The ID of the project this request is associated with.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/multicategoricalInvalidFormat/file/\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
path
string
true
The ID of the project this request is associated with.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/payoffMatrices/?offset=0&limit=100\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/payoffMatrices/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XDELETEhttps://app.datarobot.com/api/v2/projects/{projectId}/payoffMatrices/{payoffMatrixId}/\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XPUThttps://app.datarobot.com/api/v2/projects/{projectId}/payoffMatrices/{payoffMatrixId}/\-H"Content-Type: application/json"\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
POST /api/v2/projects/{projectId}/predictionExplanations/¶
Create a new PredictionExplanations object (and its accompanying PredictionExplanationsRecord).
In order to successfully create PredictionExplanations for a particular model and dataset, you must first
- Compute feature impact for the model via POST /api/v2/projects/{projectId}/models/{modelId}/featureImpact/
- Compute a PredictionExplanationsInitialization for the model via POST /api/v2/projects/{projectId}/models/{modelId}/predictionExplanationsInitialization/
- Compute predictions for the model and dataset via POST /api/v2/projects/{projectId}/predictions/thresholdHigh and thresholdLow are optional filters applied to speed up computation. When at least one is specified, only the selected outlier rows will have prediction explanations computed. Rows are considered to be outliers if their predicted value (in case of regression projects) or probability of being the positive class (in case of classification projects) isless than thresholdLow or greater than thresholdHigh. If neither is specified, prediction explanations will be computed for all rows.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/predictionExplanations/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/predictionExplanations/{predictionExplanationsId}/¶
Retrieve stored Prediction Explanations.
Each PredictionExplanationsRow retrieved corresponds to a row of the prediction dataset, although some rows may not have had prediction explanations computed depending on the thresholds selected.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/predictionExplanations/{predictionExplanationsId}/?limit=3000\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
This many results will be skipped.
limit
query
integer
true
At most this many results are returned. The default may change and a new maximum limit may be imposed without notice.
excludeAdjustedPredictions
query
string
false
Whether to include adjusted prediction in PredictionExplanationsRow response.
projectId
path
string
true
The project ID
predictionExplanationsId
path
string
true
The ID of the PredictionExplanationsRecord to retrieve.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/predictionExplanationsRecords/¶
List PredictionExplanationsRecord objects for a project.
These contain metadata about the computed prediction explanations and the location at which the PredictionExplanations can be retrieved.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/predictionExplanationsRecords/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
This many results will be skipped.
limit
query
integer
false
At most this many results are returned.
modelId
query
string
false
If specified, only prediction explanations records computed for this model will be returned.
# You can also use wget
curl-XDELETEhttps://app.datarobot.com/api/v2/projects/{projectId}/predictionExplanationsRecords/{predictionExplanationsId}/\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
projectId
path
string
true
The project ID
predictionExplanationsId
path
string
true
The ID of the PredictionExplanationsRecord to retrieve.
To perform this operation, you must be authenticated by means of one of the following methods:
BearerAuth
GET /api/v2/projects/{projectId}/predictionExplanationsRecords/{predictionExplanationsId}/¶
Retrieve a PredictionExplanationsRecord object.
A PredictionExplanationsRecord contains metadata about the computed prediction explanations and the location at which the PredictionExplanations can be retrieved.
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/predictionExplanationsRecords/{predictionExplanationsId}/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/shapMatrices/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
Parameters
Name
In
Type
Required
Description
offset
query
integer
false
Number of results to skip.
limit
query
integer
false
At most this many results are returned. The default may change without notice.
# You can also use wget
curl-XPOSThttps://app.datarobot.com/api/v2/projects/{projectId}/shapMatrices/\-H"Content-Type: application/json"\-H"Authorization: Bearer {access-token}"
# You can also use wget
curl-XGEThttps://app.datarobot.com/api/v2/projects/{projectId}/shapMatrices/{shapMatrixId}/\-H"Accept: application/json"\-H"Authorization: Bearer {access-token}"
The average actual value of the target in the bin. null if there are no entries in the bin or if this is an anomaly detection project.
endDate
string(date-time)
true
The datetime of the end of the bin (exclusive).
frequency
number¦null
true
As indicated by the frequencyType in the Metadata, used to determine what the averages mentioned above are taken over. null if there are no entries in the bin.
predicted
number¦null
true
The average prediction of the model in the bin. null if there are no entries in the bin.
The Durbin-Watson statistic for the chart data. Value is between 0 and 4. Returns -1 when the statistic is invalid for the data, e.g. if this is an anomaly detection project.
The Durbin-Watson statistic for the chart data. Value is between 0 and 4. Durbin-Watson statistic is a test statistic used to detect the presence of autocorrelation at lag 1 in the residuals (prediction errors) from a regression analysis. More info https://wikipedia.org/wiki/Durbin%E2%80%93Watson_statistic
number of times this class is seen in the validation data
className
string
true
name of the class
confusionMatrixOneVsAll
[array]
true
2d array representing 2x2 one vs all matrix. This represents the True/False Negative/Positive rates as integer for each class. The data structure looks like: [ [ True Negative, False Positive ], [ False Negative, True Positive ] ]
f1
number
true
F1 score
precision
number
true
precision score
predictedCount
integer
true
number of times this class has been predicted for the validation data
per-class information including one vs all scores in a format specified below
classes
[string]
true
class labels from the dataset, union of row classes & column classes. This field is deprecated as of v2.13. The rows and columns may have different class labels when using query parameters to retrieve a slice of the matrix; please use 'rowClasses' and 'colClasses' instead.
colClasses
[string]
true
class labels on columns of confusion matrix
confusionMatrix
[array]
true
2d array of ints representing confusion matrix, aligned with rowClasses and 'colClasses'array. For confusionMatrix[A][B] we can get an integer that represents the number of times 'if class with index A was correct we have class with index B predicted' (if the orientation is 'actual').
list of all class names in the full confusion matrix, sorted by the orderBy parameter.
datasetId
string
true
The dataset to retrieve a Confusion chart from.
modelId
string
true
The model to retrieve a Confusion chart from.
projectId
string
true
The project to retrieve a Confusion chart from.
relevantClassesPositions
[array]
true
Matrix to highlight important cell blocks in the confusion chart. Intended to represent a thumbnail view, where the relevantClassesPositions array has a 1 in thumbnail cells that are of interest, and 0 otherwise. The dimensions of the implied thumbnail will not match those of the confusion matrix, e.g. a twenty-class confusion matrix may have a 2x2 thumbnail.
totalMatrixSum
integer
true
Sum of all values in the full confusion matrix (equal to the number of points considered).
Either null or an array of up to 10 ShapleyFeatureContribution objects. Only rows with the highest anomaly scores have Shapley explanations calculated.
Compute plots for a specific backtest (use the backtest index starting from zero) or holdout. If not specified the first backtest (backtest index 0) will be used.
oneOf
Name
Type
Required
Restrictions
Description
» anonymous
integer
false
maximum: 19 minimum: 0
none
xor
Name
Type
Required
Restrictions
Description
» anonymous
string
false
none
continued
Name
Type
Required
Restrictions
Description
forecastDistanceEnd
integer
false
minimum: 0
The end of forecast distance range (forecast window) to compute. If not specified, the last forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
forecastDistanceStart
integer
false
minimum: 0
The start of forecast distance range (forecast window) to compute. If not specified, the first forecast distance for this project will be used. Forecast distance specifies the number of time steps between the predicted point and the origin point. Only available for time series supervised projects.
fullAverage
boolean
false
Whether to compute an average plot for all series. Only available for time series multiseries projects.
seriesIds
[string]
false
maxItems: 1000 minItems: 1
Only available for time series multiseries projects. Each element should be a name of a single series in a multiseries project. It is possible to compute a maximum of 1000 series per one request. If not specified the first 1000 series in alphabetical order will be computed. It is not possible to specify fullAverage: true while also setting seriesIds. This parameter can only be specified after first 1000 series in alphabetical order are computed.
Any extended message to include about the result. For example, if a job is submitted that is a duplicate of a job that has already been added to the queue, the message will mention that no new job can be created.
Actual value column name that contains actual values to be used for computing scores and insights for unsupervised projects only. This value can be set once for a dataset and cannot be changed.
datasetId
string
true
The dataset to compute predictions for - must have previously been uploaded.
Metric to use for calculating fairness. Can be one of proportionalParity, equalParity, predictionBalance, trueFavorableAndUnfavorableRateParity or FavorableAndUnfavorablePredictiveValueParity. Used and required only if Bias & Fairness in AutoML feature is enabled.
A list which contains feature impact scores for each feature used by a model. If the model has more than 1000 features, the most important 1000 features are returned.
ranRedundancyDetection
boolean
true
Indicates whether redundant feature identification was run while calculating this feature impact.
rowCount
integer¦null
true
The number of rows that was used to calculate feature impact. For the feature impact calculated with the default logic, without specifying the rowCount, we return null here.
Operator to apply to the named operand in the dataset. The operator 'eq' mean 'equals the single specified value'. The operator 'in' means 'is one of a list of allowed values.'
values
[anyOf]
true
maxItems: 1000 minItems: 1
Values to filter the operand by with the given operator.
Average actual value of the target in the bin. null if there are no entries in the bin.
endDate
string(date-time)
true
The datetime of the end of the bin (exclusive).
error
number¦null
true
minimum: 0
Average absolute residual value of the bin. null if there are no entries in the bin.
forecasts
[number]
true
maxItems: 100
An array of average forecasts for the model for each forecast distance. Empty if there are no forecasts in the bin. Each index in the forecasts array maps to forecastDistances array index.
frequency
integer¦null
true
Indicates number of values averaged in bin in case of a resolution change.
normalizedError
number¦null
true
maximum: 1 minimum: 0
Normalized average absolute residual value of the bin. null if there are no entries in the bin.
An array of available forecast distances for the completed status. If there are no forecast distances for this status, it will not appear in the response.
errored
[integer]
false
maxItems: 1000 minItems: 1
An array of available forecast distances for the errored status. If there are no forecast distances for this status, it will not appear in the response.
inProgress
[integer]
false
maxItems: 1000 minItems: 1
An array of available forecast distances for the inProgress status. If there are no forecast distances for this status, it will not appear in the response.
insufficientData
[integer]
false
maxItems: 1000 minItems: 1
An array of available forecast distances for the insufficientData status. If there are no forecast distances for this status, it will not appear in the response.
notCompleted
[integer]
false
maxItems: 1000 minItems: 1
An array of available forecast distances for the notCompleted status. If there are no forecast distances for this status, it will not appear in the response.
Count of specified frequent value in the sample, weighted by exposure or weights
dataQuality
string
true
Any data quality issue associated with this particularvalue of the feature. Possible data quality types include 'excess_zero', 'inlier', 'disguised_missing_value', and 'no_issues_found' and the relevant statistics.
target
number¦null
true
Average target value for the specified frequent value if the target is binary or numeric. With weights or exposure, this becomes a weighted average. If the target is not set, it returns None.
value
any
true
Specified frequent value, either a float or a string, like =All Others+
The average of the actual target values for the rows in the bin.
binWeight
number
true
How much data is in the bin. For projects with weights, it is the sum of the weights of all rows in the bins; otherwise, it is the number of rows in the bin.
predicted
number
true
The average of predicted values of the target for the rows in the bin.
The average of the actual target values for the rows in the bin.
binWeight
number
true
How much data is in the bin. For projects with weights, it is the sum of the weights of all rows in the bins; otherwise, it is the number of rows in the bin.
predicted
number
true
The average of predicted values of the target for the rows in the bin.
Matrix to highlight important cell blocks in the confusion chart. Intended to represent a thumbnail view, where the relevantClassesPositions array has a 1 in thumbnail cells that are of interest, and 0 otherwise. The dimensions of the implied thumbnail will not match those of the confusion matrix, e.g. a twenty-class confusion matrix may have a 2x2 thumbnail.
source
string
true
Source of the chart.
totalMatrixSum
integer
true
minimum: 0
Sum of all values in the full confusion matrix (equal to the number of points considered)
Also known as the r-squared value. This value is calculated over the downsampled dataset, not the full input
data
[array]
true
The rows of chart data in [actual, predicted, residual, row number] form. If the row number was not available at the time of model creation, the row number will be null. NOTE: In DataRobot v5.2, the row number will not be included.
A list which contains feature impact scores for each feature used by a model. If the model has more than 1000 features, the most important 1000 features are returned.
A list of feature importance scores for each class in multiclass project.
count
integer
true
Number of feature impact records in a given batch.
next
string(uri)¦null
true
URL for the next page of results or null.
previous
string(uri)¦null
true
URL for the next page of results or null.
ranRedundancyDetection
boolean
true
Indicates whether redundant feature identification was run while calculating this feature impact. Currently always False, as redundant feature identification isn't supported for multiclass in DataRobot.
shapBased
boolean
true
Indicates whether feature impact was calculated using Shapley values. Currently always False, as SHAP isn't supported for multiclass in DataRobot.
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
Key for multilabel insights, unique per project, feature, and EDA stage. The most recent key can be retrieved via GET /api/v2/projects/{projectId}/features/ or [GET /api/v2/projects/{projectId}/features/{featurename:featureName}/][get-apiv2projectsprojectidfeaturesfeaturenamefeaturename]
Name of the protected class the score is calculated for.
entriesCount
integer
true
minimum: 0
The number of entries of the class in the analysed data.
isStatisticallySignificant
boolean
true
Flag to tell whether the score can be treated as statistically significant. In other words, whether we are confident enough with the score for this protected class.
A list which contains feature impact scores for each feature used by a model. If the model has more than 1000 features, the most important 1000 features are returned.
next
string(uri)¦null
true
URL for the next page of results or null.
previous
string(uri)¦null
true
URL for the next page of results or null.
ranRedundancyDetection
boolean
true
Indicates whether redundant feature identification was run while calculating this feature impact.
rowCount
integer¦null
true
The number of rows that was used to calculate feature impact. For the feature impact calculated with the default logic, without specifying the rowCount, we return null here.
shapBased
boolean
true
Indicates whether feature impact was calculated using Shapley values. True for anomaly detection models when the project is unsupervised, as permutation approach is not applicable. Note that supervised projects must use an alternative route for SHAP impact: /api/v2/projects/(projectId)/models/(modelId)/shapImpact/
The name of the feature contributing to the prediction.
featureValue
string
true
The value the feature took on for this row. For image features, this value is the URL of the input image (New in v2.21).
imageExplanationUrl
string¦null
true
For image features, the URL of the image containing the input image overlaid by the activation heatmap. For non-image features, this field is null.
label
string
true
Describes what this model output corresponds to. For regression projects, it is the name of the target feature. For classification projects, it is a level from the target feature. For Anomaly Detection models it is an Anomaly Score.
List of class names that will be explained for each row for multiclass. Mutually exclusive with numTopClasses. If neither specified - we assume numTopClasses=1.
datasetId
string
true
The dataset ID.
maxExplanations
integer
false
maximum: 10 minimum: 0
The maximum number of prediction explanations to supply per row of the dataset.
modelId
string
true
The model ID.
numTopClasses
integer
false
maximum: 10 minimum: 1
Number of top predicted classes for each row that will be explained for multiclass. Mutually exclusive with classNames. If neither specified - we assume numTopClasses=1.
thresholdHigh
number¦null
false
The high threshold, above which a prediction must score in order for prediction explanations to be computed. If neither thresholdHigh nor thresholdLow is specified, prediction explanations will be computed for all rows.
thresholdLow
number¦null
false
The lower threshold, below which a prediction must score in order for prediction explanations to be computed for a row in the dataset. If neither thresholdHigh nor thresholdLow is specified, prediction explanations will be computed for all rows.
The maximum number of prediction explanations to supply per row of the dataset.
thresholdHigh
number¦null
false
The high threshold, above which a prediction must score in order for prediction explanations to be computed. If neither thresholdHigh nor thresholdLow is specified, prediction explanations will be computed for all rows.
thresholdLow
number¦null
false
The lower threshold, below which a prediction must score in order for prediction explanations to be computed for a row in the dataset. If neither thresholdHigh nor thresholdLow is specified, prediction explanations will be computed for all rows.
Each is a PredictionExplanationsRow. They represent a small sample of prediction explanations that could be generated for a particular dataset. They will have the same schema as the data array in the response from GET /api/v2/projects/{projectId}/predictionExplanations/{predictionExplanationsId}/. As of v2.21 only difference is that there is no forecastPoint in response for time series projects.
Describes what this model output corresponds to. For regression projects, it is the name of the target feature. For classification projects, it is a level from the target feature. For Anomaly Detection models it is an Anomaly Score.
value
number
true
The output of the prediction. For regression projects, it is the predicted value of the target. For classification projects, it is the predicted probability the row belongs to the class identified by the label.
Timestamp referencing when computation for these prediction explanations finished.
id
string
true
The PredictionExplanationsRecord ID.
maxExplanations
integer
true
The maximum number of codes generated per prediction.
modelId
string
true
The model ID.
numColumns
integer
true
The number of columns prediction explanations were computed for.
predictionExplanationsLocation
string
true
Where to retrieve the prediction explanations.
predictionThreshold
number¦null
true
The threshold value used for binary classification prediction.
projectId
string
true
The project ID.
thresholdHigh
number¦null
true
The prediction explanation high threshold. Predictions must be above this value (or below the thresholdLow value) to have PredictionExplanations computed.
thresholdLow
number¦null
true
The prediction explanation low threshold. Predictions must be below this value (or above the thresholdHigh value) to have PredictionExplanations computed.
'exposureNormalized' (for regression projects with exposure) or 'N/A' (for classification projects) The value of 'exposureNormalized' indicates that prediction outputs are adjusted (or divided) by exposure. The value of 'N/A' indicates that no adjustments are applied to the adjusted predictions and they are identical to the unadjusted predictions.
count
integer
true
How many rows of prediction explanations were returned.
Also known as the r-squared value. This value is calculated over the downsampled dataset, not the full input.
data
[array]
true
The rows of chart data in [actual, predicted, residual, row number] form. If the row number was not available at the time of model creation, the row number will be null.
Data to plot a histogram of residual values. The object contains three keys, intervalStart, intervalEnd, and occurrences, the number of times the predicted value fits within that interval. For all but the last interval, the end value is exclusive.
residualMean
number
true
The arithmetic mean of the predicted value minus the actual value over the downsampled dataset.
standardDeviation
number
true
The Standard Deviation value measures variation in the dataset. A low value indicates that the data points tend to be close to the mean; a high value indicates that the data points are spread over a wider range of values.
The backtesting score for this series. If backtesting has not been run for this model, this score will be null.
cluster
string¦null
false
The cluster associated with this series.
duration
string(duration)
true
The duration of this series formatted as an ISO 8601 duration string.
endDate
string(date-time)¦null
false
The ISO-formatted end date of this series.
holdoutScore
number¦null
false
The holdout set score for this series. If holdout is locked for the project, this score will be null.
multiseriesId
string
true
A DataRobot-generated ID corresponding to a single series in a multiseries dataset.
multiseriesValues
[string]
true
The actual values of series ID columns from the dataset.
rowCount
integer
true
The number of rows available for this series in the input dataset.
startDate
string(date-time)¦null
false
The ISO-formatted start date of this series.
targetAverage
any
false
For regression projects, this is the average (mean) value of target values for this series.For classification projects, this is the ratio of the positive class in the target for this series.
For classification - values of the target class for corresponding word or ngram. For regression - null.
coefficient
number
true
maximum: 1 minimum: -1
Describes effect of this ngram on the target. A large negative value means a strong effect toward the negative class in classification projects and a smaller predicted target value in regression projects. A large positive value means a strong effect toward the positive class and a larger predicted target value respectively.
count
integer
true
Number of rows in the training sample where this ngram appears.
frequency
number
true
maximum: 1 minimum: 0 (exclusive)
Frequency of this ngram relative to the most common ngram.
isStopword
boolean
true
True for ngrams that DataRobot evaluates as stopwords.
ngram
string
true
Word or ngram value.
variable
string
true
String representation of the ngram source - contains column name and, for some models, preprocessing details. E.g. NGRAM_OCCUR_L2_colname will be for ngram occurrences count using L2 normalization from the colname column.