labeledSensorData(featureColumns:labelColumn:recordingFileColumn:)

Declaration

func labeledSensorData(featureColumns: [String], labelColumn: String? = nil, recordingFileColumn: String? = nil) throws -> MLDataTable

Parameters

• featureColumns:

  The names of the feature columns the method includes in the Mldatatable it generates.

• labelColumn:

  The name of the label column. This parameter must not be nil if the data source uses Directorywithdataandannotation(at:annotationfilename:timestampcolumn:labelstarttimecolumn:labelendtimecolumn:).

• recordingFileColumn:

  The name of the column with the recording file names. This parameter must not be nil if the data source uses Directorywithdataandannotation(at:annotationfilename:timestampcolumn:labelstarttimecolumn:labelendtimecolumn:).

Return Value

A new MLDataTable instance.

Discussion

The labelColumn and recordingFileColumn parameters are optional if the data source is MLActivityClassifier.DataSource.labeledDirectories(at:). If nil, the method names the data table’s label column and data file column “label” and “recordingFile”, respectively.

See Also

Generating data tables from a data source

• stratifiedSplit(proportions:seed:featureColumns:labelColumn:recordingFileColumn:)