Tabulate the columns of a matrix in R-CodePudding

I have this piece of code that calculates the vector "week1" (predefined in the exercise), based on the number of times a value is repeated in that column. I would need to create a function that does the same for the other 6 columns (from 2 to 7), using the function apply, but I struggle a bit with functions yet.

This is the piece of code:

    matrix <- matrix(1, 16, 7)

    week1 <- matrix[,1]
    times <- tabulate(week1, 4)
    vectorW1 <- c(1/times[1], 1-1/times[1], 1/times[2], 1-1/times[2],
        1/times[3], 1-1/times[3],1/times[4], 1-1/times[4])

Thank you all in advance for your help!

CodePudding user response：

I've no idea what this code is trying to do, nor what the target function is attempting to do.

First, let us make it into a function

mat_to_weights <- function(data) {
  stopifnot(inherits(data, "matrix"))
  result <- matrix(NA, nrow = 8, ncol = ncol(data))
  for (col in seq_len(ncol(data))) {
    
    week <- data[, col]
    week
    times <- tabulate(week, 4)
    times
    vector <-
      c(
        1 / times[1],
        1 - 1 / times[1],
        1 / times[2],
        1 - 1 / times[2],
        1 / times[3],
        1 - 1 / times[3],
        1 / times[4],
        1 - 1 / times[4]
      )
    result[,col] <- vector
  }
  result
}
mat_to_weights(data = matrix)
> mat_to_weights(data = matrix)
       [,1]   [,2]   [,3]   [,4]   [,5]   [,6]   [,7]
[1,] 0.0625 0.0625 0.0625 0.0625 0.0625 0.0625 0.0625
[2,] 0.9375 0.9375 0.9375 0.9375 0.9375 0.9375 0.9375
[3,]    Inf    Inf    Inf    Inf    Inf    Inf    Inf
[4,]   -Inf   -Inf   -Inf   -Inf   -Inf   -Inf   -Inf
[5,]    Inf    Inf    Inf    Inf    Inf    Inf    Inf
[6,]   -Inf   -Inf   -Inf   -Inf   -Inf   -Inf   -Inf
[7,]    Inf    Inf    Inf    Inf    Inf    Inf    Inf
[8,]   -Inf   -Inf   -Inf   -Inf   -Inf   -Inf   -Inf

Next, instead of a matrix with 1s, let us try to use dummy matrix from pracma::magic(n).

> mat_to_weights(data = pracma::magic(1))
     [,1]
[1,]    1
[2,]    0
[3,]  Inf
[4,] -Inf
[5,]  Inf
[6,] -Inf
[7,]  Inf
[8,] -Inf
> mat_to_weights(data = pracma::magic(2))
     [,1] [,2]
[1,]    1  Inf
[2,]    0 -Inf
[3,]  Inf    1
[4,] -Inf    0
[5,]  Inf    1
[6,] -Inf    0
[7,]    1  Inf
[8,]    0 -Inf
Warning message:
In pracma::magic(2) : There is no magic square of order 2.
> mat_to_weights(data = pracma::magic(3))
     [,1] [,2] [,3]
[1,]  Inf    1  Inf
[2,] -Inf    0 -Inf
[3,]  Inf  Inf    1
[4,] -Inf -Inf    0
[5,]    1  Inf  Inf
[6,]    0 -Inf -Inf
[7,]    1  Inf  Inf
[8,]    0 -Inf -Inf
> mat_to_weights(data = pracma::magic(4))
     [,1] [,2] [,3] [,4]
[1,]  Inf  Inf  Inf    1
[2,] -Inf -Inf -Inf    0
[3,]  Inf    1  Inf  Inf
[4,] -Inf    0 -Inf -Inf
[5,]  Inf  Inf    1  Inf
[6,] -Inf -Inf    0 -Inf
[7,]    1  Inf  Inf  Inf
[8,]    0 -Inf -Inf -Inf
> mat_to_weights(data = pracma::magic(5))
     [,1] [,2] [,3] [,4] [,5]
[1,]  Inf  Inf    1  Inf  Inf
[2,] -Inf -Inf    0 -Inf -Inf
[3,]  Inf  Inf  Inf    1  Inf
[4,] -Inf -Inf -Inf    0 -Inf
[5,]  Inf  Inf  Inf  Inf    1
[6,] -Inf -Inf -Inf -Inf    0
[7,]    1  Inf  Inf  Inf  Inf
[8,]    0 -Inf -Inf -Inf -Inf

Again, not clear this is supposed to do, but the function works.

Let us rewrite it with apply in mind.


mat_to_weights <- function(data) {
  stopifnot(inherits(data, "matrix"))
  result <- matrix(NA, nrow = 8, ncol = ncol(data))
  apply(data, 2, function(week) {
    times <- tabulate(week, 4)
    vector <-
      c(
        1 / times[1],
        1 - 1 / times[1],
        1 / times[2],
        1 - 1 / times[2],
        1 / times[3],
        1 - 1 / times[3],
        1 / times[4],
        1 - 1 / times[4]
      )
    vector
  })
}

Other implementations to consider are of course sweep, but for now this works..