GMMPredict Arguments - Teradata Vantage

OutputFormat

[Optional] Specify how the function outputs the weights that it assigns to each of the top N clusters:

Option	Description
'sparse' (Default)	Function outputs each weight to separate output table row.
'dense'	Function outputs all weights to single output table row.

TopK

[Optional] Specify the number of cluster weights that the function outputs. This value must be an INTEGER. For the value n, the function outputs for each data point the cluster with the greatest weight, the cluster with the second-greatest weight, and so on, ending with the cluster with the kth-greatest weight.

Default: 1

OutputLogLikelihood

[Optional] Specify whether to output the log likelihood of an observation, given the data.

Default: 'false'

Accumulate

[Optional] Specify the names of testdata columns to copy to the output table.

TargetColumns

[Optional] Specify the names of testdata columns that correspond to the attributes in the modeldata table.

Default: All testdata columns except the first

IDColumn

[Optional] Specify the input table column that defines the row identifier.

Default: First input table column