CODEWITHSUNDEEP
r
Parseltongue
Parseltongue

Reputation: 11657

Divide every row in a matrix by corresponding row in vector

Imagine I have the following matrix:

mat <- read.table(text="
   0 25  7 25  9 20 26 26 22 19 22 25
  25  0 21 43 23 37 43 43 41 33 36 46
   7 21  0 21 22 20 22 37 35 32 35 45
  25 43 21  0 26 35 43 43 38 33 37 49
   9 23 22 26  0 21 24 23 24 19 22 31
  20 37 20 35 21  0 34 34 39 39 43 51
  26 43 22 43 24 34  0 44 36 33 37 48
  26 43 37 43 23 34 44  0 50 47 51 77
  22 41 35 38 24 39 36 50  0 47 50 78
  19 33 32 33 19 39 33 47 47  0 73 79
  22 36 35 37 22 43 37 51 50 73  0 87
  25 46 45 49 31 51 48 77 78 79 87  0           
", header=F)

mat <- as.matrix(mat)

I want to compute a specific network density measure which is defined as: enter image description here

where Z_ijk is the number of co-occurrences between team members i and j , and max(Z_ijk) is the maximum number of co-occurrences that team member i has with any other member of team k. N_k is the number of members on the team.

To get the maximum value in each row, I did the following:

max_values <- apply(mat, 1, max)

> max_values
 1  2  3  4  5  6  7  8  9 10 11 12 
26 46 45 49 31 51 48 77 78 79 87 87

I thought I could divide the matrix by max_values using sweep:

sweep(mat, 1, max_values, FUN = '/')

But it does not yield the intended output. Any thoughts?

The expected output would be: divide every value in row 1 by 26, every value in row 2 by 46, and so on.

Upvotes: 1

Views: 113

Answers (1)

akrun
akrun

Reputation: 887108

We can do this by dividing with rowMaxs (from matrixStats)

library(matrixStats)
mat/rowMaxs(mat)

Upvotes: 2

Related Questions