ListRocCurves: Retrieve ROC curve data for a model for all available data...

In datarobot: 'DataRobot' Predictive Modeling API

Retrieve ROC curve data for a model for all available data partitions (see DataPartition)

Description

Retrieve ROC curve data for a model for all available data partitions (see DataPartition)

Usage

ListRocCurves(model, fallbackToParentInsights = FALSE)

Arguments

model	dataRobotModel. A DataRobot model object like that returned by GetModel.
fallbackToParentInsights	logical. If TRUE, this will return the lift chart data for the model's parent if the lift chart is not available for the model and the model has a parent model.

Value

list of lists where each list is renamed as the data partitions source and returns the following components:

• source. Character: data partitions for which ROC curve data is returned (see DataPartition).

• negativeClassPredictions. Numeric: example predictions for the negative class for each data partition source.

• rocPoints. data.frame: each row represents pre-calculated metrics (accuracy, f1_score, false_negative_score, true_negative_score, true_positive_score, false_positive_score, true_negative_rate, false_positive_rate, true_positive_rate, matthews_correlation_coefficient, positive_predictive_value, negative_predictive_value, threshold) associated with different thresholds for the ROC curve.

• positiveClassPredictions. Numeric: example predictions for the positive class for each data partition source.

Examples

## Not run: 
  projectId <- "59a5af20c80891534e3c2bde"
  modelId <- "5996f820af07fc605e81ead4"
  model <- GetModel(projectId, modelId)
  ListRocCurves(model)

## End(Not run)

datarobot documentation built on May 29, 2024, 4:36 a.m.