# Fisher's exact test with more than 2x2 in R or python?

So I have the following in R:

A=c(0,0,4,0,0,0,0,0,3,3,0,0,0,0,0,0,0,0,10,0,0)
B=c(5,0,53,1,18,9,50,2,67,27,4,5,3,0,38,0,3,1,94,0,0)

tab=as.table(rbind(A,B))
row.names(tab)=c('responders','non_responders')
fisher.test(tab, workspace = 2e8)


Basically it is a measurement of 21 variables of A & B. I don't really understand the fisher.test documentation on the R website, but with the way that I ran it, is it really telling me the difference between A & B distribution of variables?

If there is an alternative appropriate statistical test that will tell me if there is a difference between A & B distributions of measured variables then please let me know as well.

There is a good reason to keep the instances where there are zeros in the same column in the table as well. This is just a single dataset. I have approximately 300 like this where there is a measurement in each column at least once. And the goal is find which of the 300 are statistically different.

• You should include your output in the question! Not everyone capable of answering the question will have the ability or inclination to run your code. Do you understand how to interpret the p-value it returns ($p\approx 0.41$)? – Glen_b Jul 3 '20 at 8:27
• (I ask the last question because it will impact whether an answer needs to explain that as well.) – Glen_b Jul 3 '20 at 10:14