1.1 - 8.10 - CFilter Syntax Elements - Teradata Vantage

Teradata Vantage™ - Machine Learning Engine Analytic Function Reference

Teradata Vantage
Release Number
October 2019
Content Type
Programming Reference
Publication ID
English (United States)
Specify the name of the output table that the function creates. The table must not exist.
Specify the names of the InputTable columns that contain the data to filter.
Specify the names of join columns, which the function uses as follows:
  1. The function uses the items in each join column to define groups of items listed in the input columns.
  2. The function tries to identify items in each input column that often appear in the same group.

For example, a join column might contain a list of sales transactions from a store, and the input column might contain each individual item purchased at the store. A sales transaction can include multiple items. For each sales transaction, the function tries to identify items that often appear in the same sales transaction (that is, items that are often purchased together).

[Optional] Specify the names of the input columns to copy to the output table. The function partitions the input data and the output table on these columns.
Specifying a column as both an partition_column and a join_column causes incorrect counts in partitions.
This syntax element makes the function output nondeterministic unless each partition_column is unique in the group defined by JoinColumns (for more information, see Nondeterministic Results and UniqueID Syntax Element).
Default behavior: The function treats the input data as belonging to one partition.
[Optional] Specify the name of the output column to use as the partition key.
Default: 'col1_item1'
[Optional] Specify the maximum size of the item set.
The function uses max_item_set to determine the size of the data structures it uses to accumulate intermediate results. If the number of distinct items in an target_column is greater than max_item_set, the function might report incorrect results without an error message.
Default: 100