Approximate Cardinality Example - Teradata Vantage

This example calculates the number of distinct values for each specified column with a 1% error rate (accuracy).

Input

The input table, cracker, has more than 3000 rows of price and advertisement information for U.S. cracker brands Sunshine, Keebler, Nabisco, and a private label (such as a store brand). Input column names have these meanings:

SQL Call

SELECT * FROM ApproxCardinalityReduce (
  ON (
    SELECT * FROM ApproxCardinalityMap (
	  ON cracker AS InputTable
	  USING
	  InputColumns ('sn','id',
        'dispsunshine','dispkeebler','dispnabisco','dispprivate',
        'featsunshine','featkeebler','featnabisco','featprivate',
        'pricesunshine','pricekeebler','pricenabisco','priceprivate',
        'choice')
	  ErrorRate (1)
	) as dt1
  ) PARTITION BY column_name
) AS dt ;

Output