Fisher's exact test of 2x3 contingency tables permits calculation of precise probabilities in situation where, as a consequence of small cell frequencies, the much more rapid normal approximation and chi-square calculations are liable to be inaccurate. The Fisher's exact test involves the computations of several factorials to obtain the probability of the observed and each of the more extreme tables. Factorials growth quickly, so it's necessary use logarithms of factorials. This computations is very easy in Matlab because x!=gamma(x+1) and log(x!)=gammaln(x+1). This
function is now fully vectorized to speed up the computation.
For negative values, gammaln would return INF so that table is not added to computation. Maybe in previous version of gammaln this is not allowed and it calls error. So rows in Tables matrix with negative values must be erased before to give it to gammaln.
[I J]=find(Tables<0); Tables(I,:)=[];
will do this
I believe I am experiencing the same error and crashing of the script as Matthew had previously described. Certain input matrices cause the program to crash. My input was shown below. You had previously responded that your trial of his input did not crash; however, I am experiencing the same bug. Is it possible that you have a different version on your own computer?
myfisher23([19 13 15; 10 8 17])
Error using gammaln
Input must be nonnegative.
Error in myfisher23 (line 108)
zf=gammaln(Tables+1); %compute log(x!)