Permutation Method to compare paired data not normally distributed
This script is usefull if the user has to handle with paired dataset. It is not a simple Non-parametric test because do not consider outliers. The procedure is similar of a paired-t-test but with 15000 permutations.
The dataset do not follow a “T-Normal” distribution
Input Two Matrices
: Control and Treatment
Return
: a T observed and a T permuted
It is possible to select the treshold by which the T-values observed are significant or not to the respect of the T-values permuted.
In Script.r it is possible to find an example on how to use the script.