Denorm - INPUT - Analysis Parameters - Teradata Warehouse Miner

Teradata Warehouse Miner User Guide - Volume 2ADS Generation

Product
Teradata Warehouse Miner
Release Number
5.4.4
Published
July 2017
Language
English (United States)
Last Update
2018-05-03
dita:mapPath
fcf1492702067123.ditamap
dita:ditavalPath
ft:empty
dita:id
B035-2301
Product Category
Software
  1. On the Denorm dialog box, click on INPUT.
  2. Click on analysis parameters.
    Denorm > Input > Analysis Parameters

  3. On this screen, select:
    • Values to Denormalize — A list of values and prefixes which are valid values in the column specified in Index Remove Columns. Use the Add and Remove buttons to set values for the following:
      • Prefix — An optional string (must be a valid Teradata word) that will define the unique Value specified.
      • Value in <remove column> — A list of distinct values which the column specified in Index Remove Columns takes on.

    • Add button — Both Prefix and Value in <remove column> can be specified manually by clicking on the Add button and typing the required values.
    • Remove button — Remove the currently highlighted Prefix and Value in <remove column>.

    • Values — Brings up the Denorm Values Wizard dialog box.
      Denorm Values Wizard

      A status message indicating which column values are being fetched appears. Once this is complete, the distinct values of the column being removed are listed in the left-most column. These values can be dragged and dropped into the right-most column, or selected via the Add button. Similarly, they can be removed via the Remove button. Once the values to be denormalized are moved to Selected values, click Finish to return to the Teradata Warehouse Miner user interface or to continue to select the values of the next Denormalize Column, if there is more than one.

      When the Values load process is finished, a default value is generated for each Column Prefix by concatenating the values of the Index Remove Columns, each value followed by an underscore character. If the combination of the prefix and the longest Denormalize Column name or alias will be greater than 30 characters in length, the prefix is left blank, to be filled in by the user. Note that if the name of a Denormalize Column or the values of the Index Remove Columns are long, it may be necessary to specify a comparatively short alias for the Denormalize Column so that automatic prefixes can be generated. Otherwise, it may be necessary to specify a short prefix manually.

    • Aggregation Method — Specify an aggregation method in the case where new columns have multiple values to choose from. Valid user specified aggregation methods include MIN, MAX, AVG, SUM and COUNT. These should only be used when there are non-unique indices or when a part of the index is being ignored, that is, when part of the key is neither being retained nor removed (i.e., denormalized by).
    • Treat Undefined Index Values As — Specify ZERO instead of NULL for the value of those denormalized columns for which the value of the removed column is not the target value. The default is NULL. If DATE or TIMESTAMP columns are specified in Denormalize Columns, and ZERO is selected, an error occurs.
    • Compress undefined index values in output table — Request data compression of either NULL or ZERO values (depending on the Treat Undefined Index Values As option above) for those denormalized columns for which the value of the removed column is not the target value.