7.00.02 - Single_Tree_Predict Arguments - Aster Analytics

Teradata Aster® Analytics Foundation User GuideUpdate 2

Product
Aster Analytics
Release Number
7.00.02
Release Date
September 2017
Content Type
Programming Reference
User Guide
Publication ID
B700-1022-700K
Language
English (United States)
AttrTableGroupByColumns
Specifies the names of the columns on which attribute_table is partitioned. Each partition contains one attribute of the input data.
AttrTablePIDColumns
Specifies the names of the columns that define the data point identifiers.
AttrTableValColumn
Specifies the name of the column that contains the input values.
OutputResponseProbDist
[Optional] Specifies whether to output probabilities. If this value is true, you must specify the model on which to base the output probabilities, either in this argument or in the ModelTableName argument. The model_table is the name of a previously built model table. Default: 'false'.
For correct results, the values in the column prob_label_order in model_table must exactly match those in the prob_label_order in the model_table specified in ON { table | view | query } AS model DIMENSION. Teradata recommends that you specify the same table in both arguments.
ModelTableName
[Required if OutputResponseProbDist is true and OutputResponseProbDist does not specify model_table, otherwise disallowed] Specifies the name of the previously built model table on which to base the output probabilities.
For correct results, the values in the column prob_label_order in model_table must exactly match those in the prob_label_order in the model_table specified in ON { table | view | query } AS model DIMENSION. Teradata recommends that you specify the same table in both arguments.
Accumulate
[Optional] Specifies the names of the input columns to copy to the output table.
If you are using this function to create input for ROC, this argument must specify actual_label. See Example 2: Creating Input for ROC.