This method returns a DM_ClasBldTask value determined by the given DM_MiningData values for training and validation, and the given DM_ClasSettings value.
Syntax
Method syntax
>>-DM_ClasBldTask()..DM_defClasBldTask--(--trainData--,--validationData--,--clasSettings--)-><
Function syntax
Parameters
- trainData
- A value of type DM_MiningData containing the input source table definitions used for a Classification training run
- validationData
- A value of type DM_MiningData containing the input source table definitions used for a Classification test run
- clasSettings
- A value of type DM_ClasSettings containing the Classification settings
Return value
The DM_defClasBldTask method constructs a value
of type DM_ClasBldTask with
the given DM_MiningData and DM_ClasSettings values.
If DM_defClasBldTask is called on an existing DM_ClasBldTask value,
content is overwritten.
- If settings is NULL, this raises an exception. The exception states that clasSettings is NULL.
- If trainData is NULL, this raises an exception. The exception states that the training data value is NULL.
- If no logical data specification is defined in clasSettings, this raises an exception. The exception states that the logical data specification is missing.
- If trainData or validationData is incompatible with the logical data specification in clasSettings, this raises an exception. The exception states that the mining data and the logical data specification are not compatible.
- If validationData is NULL, no validation of the model is done during the training run.
- If a target field is not present in clasSettings, this raises an exception. The exception states that no target field has been defined.
- Otherwise, the return value is a DM_ClasBldTask value determined by the given parameters.