Converting dataframe to a matrix

Question

I have a table of species combinations that overlap geographically.

Sp_a    Sp_b   prop_overlap
Cat     Dog    1

But I want to create a matrix that looks like this:

      Cat 
Dog    1

I have approximately 180000 combinations. Is there any quick way of converting this information in R?

I have tried:

    m <- matrix( NA, ncol = max(species_int$sp_a) , nrow = max(species_int$sp_b) ) 
m[ cbind( species_int$sp_b , species_int$sp_a ) ] <- species_int$prop_overlap

Thanks

You just need to use `reshape2::acast`, like `acast(species_int, Sp_a, Sp_b)`. — Anonymous coward, Oct 31 '18 at 22:06

Michelle Ward · Answer 1 · 2018-11-01T00:44:39.203

0

Thanks, just had to rewrite it a little:

library(reshape2)
m <- acast(species_int, sp_a~sp_b, value.var = "prop_overlap")

edited Nov 01 '18 at 00:44

answered Oct 31 '18 at 22:23

Michelle Ward

61
8

If you want your answers to be understandable, then you should inlcude the library calls to load functions needed. – IRTFM Oct 31 '18 at 22:36
thanks, I have edited the above answer. – Michelle Ward Nov 01 '18 at 00:45

Converting dataframe to a matrix

1 Answers1