CODEWITHSUNDEEP
rmatrixconditional-statements
striatum
striatum

Reputation: 1598

R: Conditional deleting rows and columns in a matrix

I have a square matrix of co-occurrence type of data, like:

m <- matrix(c(30, 30, 30, 30, 20, 0, 0,
    30, 373, 30, 204, 207, 0, 290,
    30, 30, 65, 65, 20, 35, 0,
    30, 204, 65, 239, 38, 35, 156,
    20, 207, 20, 38, 207, 0, 134,
    0, 0, 35, 35, 0, 35, 0,
    0, 290, 0, 156, 134, 0, 290),
    nrow=7, byrow=TRUE)

By comparing upper-triangular + diagonal elements, there are some off-diagonals equal to diagonal. I want to remove rows and columns by satisfying:

if ((m[i,j] == m[i,i]) & (m[i,j] < m[j,j]))

Thus, leaving only the row/column that has larger occurrence, and to take out the row/column when an element always co-occur with another.

The output should be:

373  204
204  239

Thanks!

Upvotes: 2

Views: 2211

Answers (1)

flodel
flodel

Reputation: 89057

Here is a vectorized approach:

i <- as.vector(row(m))
j <- as.vector(col(m))
k <- matrix(m == m[cbind(i, i)] & m < m[cbind(j, j)], nrow(m))

#       [,1]  [,2]  [,3]  [,4]  [,5]  [,6]  [,7]
# [1,] FALSE  TRUE  TRUE  TRUE FALSE FALSE FALSE
# [2,] FALSE FALSE FALSE FALSE FALSE FALSE FALSE
# [3,] FALSE FALSE FALSE  TRUE FALSE FALSE FALSE
# [4,] FALSE FALSE FALSE FALSE FALSE FALSE FALSE
# [5,] FALSE  TRUE FALSE FALSE FALSE FALSE FALSE
# [6,] FALSE FALSE  TRUE  TRUE FALSE FALSE FALSE
# [7,] FALSE  TRUE FALSE FALSE FALSE FALSE FALSE

delete.idx <- sort(unique(i[k]))
# [1] 1 3 5 6 7

keep.idx <- setdiff(seq_len(nrow(m)), delete.idx)
# [1] 2 4

m[keep.idx, keep.idx]
#      [,1] [,2]
# [1,]  373  204
# [2,]  204  239

Upvotes: 2

Related Questions