The ML.NGRAMS function

This document describes theML.NGRAMS function, which lets you createn-grams of the input values.

You can use this function with models that supportmanual feature preprocessing. For moreinformation, see the following documents:

• End-to-end user journeys for ML models
• Contribution analysis user journey).

Syntax

ML.NGRAMS(array_input, range [, separator])

Arguments

ML.NGRAMS takes the following arguments:

• array_input: anARRAY<STRING> value that represent the tokens to bemerged.
• range: anARRAY of twoINT64 elements or a singleINT64 value. Ifyou specify anARRAY value, theINT64 elements provide the rangeof n-gram sizes to return. Provide the numerical values in order, lower tohigher. If you specify a singleINT64 value ofx, therange of n-gram sizes to return is[x, x].
• separator: aSTRING value that specifies the separator toconnect two adjacent tokens in the output. The default value iswhitespace.

Output

ML.NGRAMS returns anARRAY<STRING> value that contain the n-grams.

Example

The following example outputs all possible 2-token and 3-token combinationsfor a set of three input strings:

SELECTML.NGRAMS(['a','b','c'],[2,3],'#')ASoutput;

The output looks similar to the following:

+-----------------------+|        output         |+-----------------------+| ["a#b","a#b#c","b#c"] |+-----------------------+

What's next

• For information about feature preprocessing, seeFeature preprocessing overview.