Applying Dimension Values - Teradata Warehouse Miner

Teradata® Warehouse Miner™ User Guide - Volume 2ADS Generation

Teradata Warehouse Miner
Release Number
November 2018
English (United States)
Last Update
Product Category

Consider the following example of defining dimension values based on a column called tran_code in the input table twm_credit_tran from which a variable is being defined based on another input column called tran_amt. The valid values of tran_code may be extracted directly from the twm_credit_tran table using the Values button on the Variable Creation INPUT screen. They could also be taken from the output of a previous run of the Data Explorer or Frequency analyses. At this point, the user might select the tran_code values 'CA’, ‘CG’, and ‘PM’ as dimension values, and the combination of ‘CA’ and ‘CG’ as a fourth dimension value. A name is given to each of these dimension values to be used in conjunction with variable names in naming any variables dimensioned by this dimension value. A descriptive string may also be associated with each of the dimension values.

The dimension information is shown below for conceptual purposes in the form of two tables. Note that the Dimension Values table targets the dimension values of tran_code in a particular table. Notice that the conditions comprising the elements of the dimension may overlap. That is, they do not need to be mutually exclusive in value.

Dimension Values
Dimension Value Name Full Description
tran_code = ‘CA’ tran_code_CA Cash advance
tran_code = ‘CG’ tran_code_CG Charge
tran_code = ‘PM’ tran_code_PM Payment
tran_code IN (‘CA’,‘CG’) tran_code_CA_CG Advance or charge

Suppose the above dimension values are applied to a new variable, AVG (tran_amt), with abbreviation Amt. The select list items for the AVG (tran_amt) dimensioned by these dimension values would produce 4 variables.

Dimension Values > 4 Variables
Variable Name Full Description
tran_code_CA_Amt Average Tran Amount for Cash advances
tran_code_CG_Amt Average Tran Amount for Charges
tran_code_PM_Amt Average Tran Amount for Payments
tran_code_CA_CG_Amt Average Tran Amount for Advances or Charges

Conditions other than simple inclusion can be used in defining dimensions. In fact, any SQL construct listed previously with the exception of an Aggregation or Ordered Analytic function can be used.