Description: kstest_2s_2d.m: function to do two-sample two-dimensional Kolmogorov-Smirnov test
It s similar to the MATLAB built-in funtion "kstest2" from Statistics Toolbox except that "kstest_2s_2d" deals with two-dimensional data.
[H, pValue, KSstatistic] = kstest_2s_2d(x1, x2, alpha, tail)
The function takes in 2-column inputs "x1" and "x2", where each column represents one dimension. The other options and the outputs are of the same meaning and format as "kstest2".
Algorithm summary (Peak, 1983): consider the four quadrants (x<X,y<Y), (x<X,y>Y), (x>X,y<Y) and (x>X,y>Y) in turn, and adopt the largest of the four differences between the two empirical cumulative distributions as the final KS statistic.
test_kstest_2s_2d.m: test-bench of "kstest_2s_2d"
Please feel free to contact me if you find any bugs in the code. Hope I understood and implemented the ideas in the references correctly...
Author: Qiuyan Peng @ ECE/HKUST
Date: 11th May, 2011
Refere
To Search:
File list (Check if you may need any files):
kstest_2s_2d.m
license.txt
test_kstest_2s_2d.m
Download