1.0 - 8.00 - LARPredict Arguments - Teradata Vantage

Teradata® Vantage Machine Learning Engine Analytic Function Reference

Product
Teradata Vantage
Release Number
1.0
8.00
Release Date
May 2019
Content Type
Programming Reference
Publication ID
B700-4003-098K
Language
English (United States)
CoefMode
[Optional] Specify the mode for the CoefPositions argument, one of more of the following:
Option Description
'STEP' (Default) CoefPositions argument indicates steps corresponding to steps in model output by LAR function. CoefPositions argument can include any real values in [1, k], where k is maximum step in model.
'FRACTION' CoefPositions argument indicates fractions of L1 norm of coefficients against maximum L1 norm. Maximum L1 norm is that of full OLS solution (coefficients at last step). CoefPositions argument can include any real values in [0, 1].
'NORM' CoefPositions argument indicates L1 norm of coefficients. CoefPositions argument can include any real values in [0, max L1 norm]. For maximum L1 norm, see above.
'LAMBDA' CoefPositions argument indicates maximum absolute correlations. For definition, see description of max_abs_corr in LAR Output. CoefPositions argument can include any real values.
CoefPositions
[Optional] Specify the positions of the coefficients at which to create predictions. Each coefficient is a different DOUBLE PRECISION value in the range specified by the CoefMode argument.
ResponseColumn
[Optional] Specify the name of the response column in the input table (for prediction comparison).