datarobot (version 2.18.6)

SubmitActuals: Submit actuals for processing.

Description

The actuals submitted will be used to calculate accuracy metrics. Values are not processed immediately and may take some time to propagate through deployment systems. Submission of actuals is limited to 10,000,000 actuals per hour. For time series deployments, total actuals = number of actuals * number of forecast distances. For example, submitting 10 actuals for a deployment with 50 forecast distances = 500 total actuals. For multiclass deployments, a similar calculation is made where total actuals = number of actuals * number of classes. For example, submitting 10 actuals for a deployment with 20 classes = 200 actuals.

Usage

SubmitActuals(actuals, deploymentId, batchSize = 10000)

Arguments

actuals

dataframe. Data that describes actual values. Any strings stored as factors will be coerced to characters with as.character. Allowed columns are:

• associationId string. A unique identifier used with a prediction. Max length 128 characters.

• actualValue string or numeric. The actual value of a prediction; should be numeric for deployments with regression models or string for deployments with classification model.

• wasActedOn logical. Optional. Indicates if the prediction was acted on in a way that could have affected the actual outcome.

• timestamp POSIXt. Optional. If the datetime provided does not have a timezone, we assume it is UTC.

deploymentId

character. The ID of the deployment.

batchSize

integer. Optional. The max number of actuals in each batch request. Cannot exceed 10000.

See Also

Other deployment accuracy functions: GetDeploymentAccuracyOverTime(), GetDeploymentAccuracy(), GetDeploymentAssociationId()

Examples

if (FALSE) {
  deploymentId <- "5e319d2e422fbd6b58a5edad"
  myActuals <- data.frame(associationId = c("439917"),
                          actualValue = c("True"),
                          wasActedOn = c(TRUE))
  SubmitActuals(actuals = myActuals,
                deploymentId)
}