File Exchange

image thumbnail


version 1.0 (5.55 KB) by

Single Sample Kupier Goodness-Of-Fit Hypothesis Test.



No License

Single Sample Kupier Goodness-Of-Fit Hypothesis Test. Similar to Kolmogorov-Smirnov (K-S) test, but (K-S) test tend to be most sensitive around median value of the distribution and less sensitive at the distribution tails.

Doesn't require any Toolbox.Find if the random sample X could have the hypothesized continuous cumulative distribution function F.

Null Hypothesis Ho: "The Sample is taken form a population with cumulative distribution F(x) for all x".

This is a 2-sided test. Alternative: "The population from which the sample is taken has a cumulative distribution different form F "

Comments and Ratings (3)


Can this be used for comparison of two samples (e.g. kstest2)? If not does anyone know of a function to do this?

Ronald van Elburg

Code looks correct and efficient. Although the english in the comments needs a bit of cleaning up, the whole file is well documented. As noted before the name of the test is Kuiper test, not Kupier.

Banu Gunel

It is Kuiper test, not Kupier.

MATLAB Release
MATLAB 6.5 (R13)

Inspired: plotcdfkuiper

Download apps, toolboxes, and other File Exchange content using Add-On Explorer in MATLAB.

» Watch video