v.normal computes tests of normality on vector points.


The tests that v.normal performs are indexed below. The tests that are performed are specified by giving an index, ranges of indices, or multiple thereof.
  1. Sample skewness and kurtosis
  2. Geary's a-statistic and an approximate normal transformation
  3. Extreme normal deviates
  4. D'Agostino's D-statistic
  5. Modified Kuiper V-statistic
  6. Modified Watson U^2-statistic
  7. Durbin's Exact Test (modified Kolmogorov)
  8. Modified Anderson-Darling statistic
  9. Modified Cramer-Von Mises W^2-statistic
  10. Kolmogorov-Smirnov D-statistic (modified for normality testing)
  11. Chi-Square test statistic (equal probability classes) and the number of degrees of freedom
  12. Shapiro-Wilk W Test
  13. Weisberg-Binghams W'' (similar to Shapiro-Francia's W')
  14. Royston's extension of W for large samples
  15. Kotz Separate-Families Test for Lognormality vs. Normality


v.random random n=200
v.db.addtable random col="elev double precision"
v.what.rast random rast=elevation.10m col=elev
v.normal random tests=1-3,14 col=elev
computes the sample skewness and kurtosis, Geary's a-statistic and an approximate normal transformation, extreme normal deviates, and Royston's W for the random vector points.




James Darrell McCauley <>,
when he was at: Agricultural Engineering Purdue University

Last changed: $Date$