Common Analytic Functions in Teradata ML Engine and Teradata SQL Engine - Teradata Vantage

Teradata® Vantage User Guide

Product
Teradata Vantage
Release Number
1.0
Published
January 2019
Language
English (United States)
Last Update
2020-03-11
dita:mapPath
hfp1506029122470.ditamap
dita:ditavalPath
hfp1506029122470.ditaval
dita:id
B700-4002
lifecycle
previous
Product Category
Teradata Vantage
These analytic functions are available on both Teradata ML Engine and SQL Engine:
  • Attribution
  • Forest_Predict
  • GLMPredict
  • nPath
  • NTree
  • NaivebayesPredict
  • NaiveBayesTextClassifierPredict
  • Sessionize
  • Single_Tree_Predict
  • SparseSVMPredictor
They have common syntax and generate similar results. The analytic functions in SQL Engine are expected to run with better performance because they are native to the database.
A query that starts this way calls a SQL Engine analytic function:
SELECT * FROM function_name ( ...

For information on SQL Engine analytic function syntax, see Teradata® Database Analytic Functions, B035-1206.

A query that starts this way calls a Teradata ML Engine analytic function. You must append "@coprocessor" to the coprocessor function name specified in Teradata ML Engine Analytic Function Mapping:
SELECT * FROM function_name@coprocessor ( ...

Function aliases are not defined for the functions listed above.

For information on Teradata ML Engine analytic function syntax, see Teradata® Vantage Machine Learning Engine Analytic Function Reference, B700-4003.