fisher.test all entries of 'x' must be nonnegative and finite-CodePudding

I'm helping out a friend with statistic on R while not knowing it. I'm trying to do a simple fisher.test(df) so far I've been able to select the data he wants to analyse. Below my code

from the dim function I understand that it is a 2 dimensions matrix

library(data.table)
data = fread(myFILE, header = FALSE, select = c(9,46))

dim(data)
# [1] 68  2

print(data)
> print(d)
     V9          V46
 1: Question1 ? Question2 ?
 2: 30          No
 3: 25          Yes
 4: 24          Yes
 5: 25          Yes
 6: 27          Yes
 7: 27          Yes
 8: 25          Yes
 9: 26          Yes
10: 25          Yes
11: 24          Yes
12: 26          No
13: 24          Yes
14: 27          Yes
15: 26          Yes
16: 26          Yes
17: 26          Yes
18: 27          Yes
19: 26          Yes

data = d[-1,] # to remove the 1: Question...

fisher.test(data)
Error in fisher.test(data) : 
  all entries of 'x' must be nonnegative and finite

What am I missing? Should I convert Yes/No to 1/0 ?

Thanks for your help.

CodePudding user response：

As per the documention of ?fisher.test, we can either pass x and y as two arguments. If x is a matrix of contingency table, y argument is ignored. Or else can be both factor columns. Here, we can convert the 'data' into a frequency table with table applied on the two column data.table which returns a table object (which also inherits the matrix class)

fisher.test(table(data))

    Fisher's Exact Test for Count Data

data:  table(data)
p-value = 0.2092
alternative hypothesis: two.sided