Statistics - Teradata Warehouse Miner

Teradata Profiler Plug-in User Guide

Product
Teradata Warehouse Miner
Release Number
5.4.5
Published
November 2017
Language
English (United States)
Last Update
2018-05-03
dita:mapPath
wzn1503087325188.ditamap
dita:ditavalPath
ft:empty
dita:id
B035-2304
Product Category
Software

When dealing with numeric data columns, use statistical measures to understand the characteristics and properties of each of those columns, to assess their quality, and to look for outlying values and other possible anomalies. Statistical analysis provides several statistical measures for numeric data columns.

Given a table name and the names of numeric columns, statistical analysis determines descriptive statistics for each of the columns. Univariate statistics provided include the following:
  • Count
  • Minimum
  • Maximum
  • Mean
  • Standard deviation
  • Skewness
  • Kurtosis
  • Standard error
  • Coefficient of variance
  • Variance
  • Sum
  • Uncorrected sums of squares
  • Corrected sums of squares

For DATE columns, statistics other than count, minimum, maximum, and mean are calculated by first converting to the number of days since 1900.

If non-numeric columns are selected, you receive a Teradata SQL Error 2621: "Bad character in format or data of Table Column."

The Statistical analysis is parameterized by specifying the table and columns to analyze, options unique to the Statistical analysis, and the output and expert options.