creating a correlation matrix in R

Question

I have a file that has a matrix of 500 rows (binary scores) and 120 columns. The file is a simple matrix of 0s and 1s.

>file

00010010101010
01001010100101
00101001010001
11110101001010

I am writing a function that uses a special correlation formula to find this correlation between the rows. It takes two vector rows as input fn(row1, row2). eg. row1 and row2 and calculates this special correlation.

Example

>fn(file[1,], file[2,])
>0.32

I am able to do it for two rows but how can I create a 500x500 correlation matrix for all rows. Can someone please help with this? Thanks.

flodel · Accepted Answer · 2013-09-29 23:13:40Z

5

Try this:

corr.mat <- outer(seq_len(nrow(file)), seq_len(nrow(file)),
                  Vectorize(function(i, j) fn(file[i,], file[j,])))

If this is too slow for your needs, there might be better approaches but you'll have to show what fn is supposed to do.

P.S.: file being the name of a function in R, you should avoid using it for your own variables.

edited Sep 29, 2013 at 23:13

answered Sep 29, 2013 at 22:53

flodel

89.3k21 gold badges190 silver badges225 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

CadisEtRama Over a year ago

thank you, I will try it and see what happens. fn calculates a prespecified tetrachoric correlations. So I wanted to create a 500x500 correlation matrix for all the row variables in the data frame. I will try it and see.

Collectives™ on Stack Overflow

creating a correlation matrix in R

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related