How to do a row-wise replacement of values using R?
I have a Matrix and I would like to replace some of its values using an index vector. The problem is that R automatically does a column-wise extraction of the values as opposed to a row-wise.
You will find my code and results below:
Matrix=matrix(rep(0,42),nrow=6,ncol=7,byrow=TRUE)
v=c(1,7,11,16,18)
Matrix[v]=1
Matrix
[,1] [,2] [,3] [,4] [,5] [,6] [,7]
[1,] 1 1 0 0 0 0 0
[2,] 0 0 0 0 0 0 0
[3,] 0 0 0 0 0 0 0
[4,] 0 0 1 0 0 0 0
[5,] 0 1 0 0 0 0 0
[6,] 0 0 1 0 0 0 0
What I actually want to get is the row-wise version of this meaning:
[,1] [,2] [,3] [,4] [,5] [,6] [,7]
[1,] 1 0 0 0 0 0 1
[2,] 0 0 0 1 0 0 0
[3,] 0 1 0 1 0 0 0
[4,] 0 0 0 0 0 0 0
[5,] 0 0 0 0 0 0 0
[6,] 0 0 0 0 0 0 0
>
Apparently R does a column-wise replacement of values by default.
What is the best way to obtain a row-wise replacement of the values?
Thanks!
You could recalculate the onedimensional indizes to row- and column-indices. Supposing you have calculated the row-indices in the first column of the matrix Ind and the columnindices in the second column of Ind you can do Matrix[Ind] <- 1
Matrix <- matrix(rep(0,42),nrow=6,ncol=7,byrow=TRUE)
v <- c(1,7,11,16,18)
Row <- (v-1) %/% ncol(Matrix) +1
Col <- (v-1) %% ncol(Matrix) +1
Matrix[cbind(Row,Col)] <- 1
Matrix
# [,1] [,2] [,3] [,4] [,5] [,6] [,7]
# [1,] 1 0 0 0 0 0 1
# [2,] 0 0 0 1 0 0 0
# [3,] 0 1 0 1 0 0 0
# [4,] 0 0 0 0 0 0 0
# [5,] 0 0 0 0 0 0 0
# [6,] 0 0 0 0 0 0 0
We can do
+(matrix(seq_along(Matrix) %in% v, ncol=ncol(Matrix), nrow=nrow(Matrix), byrow=TRUE))
# [,1] [,2] [,3] [,4] [,5] [,6] [,7]
#[1,] 1 0 0 0 0 0 1
#[2,] 0 0 0 1 0 0 0
#[3,] 0 1 0 1 0 0 0
#[4,] 0 0 0 0 0 0 0
#[5,] 0 0 0 0 0 0 0
#[6,] 0 0 0 0 0 0 0
You could redo your 1's to make them row-wise or you can do the following:
Matrix=matrix(rep(0,42),nrow=6,ncol=7,byrow=TRUE)
v=c(1,7,11,16,18)
Matrix<-t(Matrix)
Matrix[v]=1
Matrix<-t(Matrix)
Matrix
[,1] [,2] [,3] [,4] [,5] [,6] [,7]
[1,] 1 0 0 0 0 0 1
[2,] 0 0 0 1 0 0 0
[3,] 0 1 0 1 0 0 0
[4,] 0 0 0 0 0 0 0
[5,] 0 0 0 0 0 0 0
[6,] 0 0 0 0 0 0 0
Related
Let's say, there are two matrices:
A <- B <- diag(3)
> A
[,1] [,2] [,3]
[1,] 1 0 0
[2,] 0 1 0
[3,] 0 0 1
I want to create a new matrix AB, which consists of all the possible combinations of rows of A and B. Expected result:
> AB
[,1] [,2] [,3] [,4] [,5] [,6]
[1,] 1 0 0 1 0 0
[2,] 1 0 0 0 1 0
[3,] 1 0 0 0 0 1
[4,] 0 1 0 1 0 0
[5,] 0 1 0 0 1 0
[6,] 0 1 0 0 0 1
[7,] 0 0 1 1 0 0
[8,] 0 0 1 0 1 0
[9,] 0 0 1 0 0 1
How to do this efficiently? And can it be extended for more than two matrices?
You can use expand.grid() and take its output to index the matrix A and B,
x <- expand.grid(1:3,1:3)
cbind(A[x[,1],], B[x[,2],])
gives,
[,1] [,2] [,3] [,4] [,5] [,6]
[1,] 1 0 0 1 0 0
[2,] 0 1 0 0 1 0
[3,] 0 0 1 0 0 1
[4,] 1 0 0 1 0 0
[5,] 0 1 0 0 1 0
[6,] 0 0 1 0 0 1
[7,] 1 0 0 1 0 0
[8,] 0 1 0 0 1 0
[9,] 0 0 1 0 0 1
EDIT:
For more than two matrices, you can use a function like below,
myfun <- function(...) {
arguments <- list(...)
a <- expand.grid(lapply(arguments, function(x) 1:nrow(x)))
do.call(cbind,lapply(seq(a),function(x) { arguments[[x]][a[,x],] }))
}
out <- myfun(A,B,C)
head(out)
gives,
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10]
[1,] 1 0 0 1 0 0 1 0 0 0
[2,] 0 1 0 1 0 0 1 0 0 0
[3,] 0 0 1 1 0 0 1 0 0 0
[4,] 1 0 0 0 1 0 1 0 0 0
[5,] 0 1 0 0 1 0 1 0 0 0
[6,] 0 0 1 0 1 0 1 0 0 0
Data:
A <- B <- diag(3)
C <- diag(4)
I am trying to use the matrix() and diag() functions to create the following pattern, but with a 100 x 100 matrix rather than 5 x 5.
5 x 5 matrix:
| 0 1 0 0 0 |
| 1 0 1 0 0 |
| 0 1 0 1 0 |
| 0 0 1 0 1 |
| 0 0 0 1 0 |
In other words, I want to have two diagonals with values of 1, one to the left of the main diagonal, and one to the right of the main diagonal.
The diag() function (actually the diag<- function) can be used for assignment:
mat <- matrix( 0, 100,100)
diag(mat) <- 1
mat[1:10,1:10]
#-----------
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10]
[1,] 1 0 0 0 0 0 0 0 0 0
[2,] 0 1 0 0 0 0 0 0 0 0
[3,] 0 0 1 0 0 0 0 0 0 0
[4,] 0 0 0 1 0 0 0 0 0 0
[5,] 0 0 0 0 1 0 0 0 0 0
[6,] 0 0 0 0 0 1 0 0 0 0
[7,] 0 0 0 0 0 0 1 0 0 0
[8,] 0 0 0 0 0 0 0 1 0 0
[9,] 0 0 0 0 0 0 0 0 1 0
[10,] 0 0 0 0 0 0 0 0 0 1
You, however, want the sub-diagonal and super-diagonal to be assigned values, so use logical expressions with col and row:
mat <- matrix( 0, 100,100)
mat[row(mat)==col(mat)-1] <- 1
mat[row(mat)==col(mat)+1] <- 1
mat[1:10,1:10]
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10]
[1,] 0 1 0 0 0 0 0 0 0 0
[2,] 1 0 1 0 0 0 0 0 0 0
[3,] 0 1 0 1 0 0 0 0 0 0
[4,] 0 0 1 0 1 0 0 0 0 0
[5,] 0 0 0 1 0 1 0 0 0 0
[6,] 0 0 0 0 1 0 1 0 0 0
[7,] 0 0 0 0 0 1 0 1 0 0
[8,] 0 0 0 0 0 0 1 0 1 0
[9,] 0 0 0 0 0 0 0 1 0 1
[10,] 0 0 0 0 0 0 0 0 1 0
(This method does not depend on having a square matrix. I have a vague memory that there is a faster method that does not require using row and col. For very large objects each of those functions returns a matrix of the same dimensions as their arguments.)
For the main diagonal, the row and column indices are the same. For the other diagonals, there is a difference of 1 between the row index and column index. Generate those indices directly and assign values in those indices.
sz = 5
m = matrix(0, sz, sz)
inds1 = cbind(r = 1:(sz-1), c = 2:sz)
inds2 = cbind(r = 2:sz, c = 1:(sz-1))
m[inds1] = 1
m[inds2] = 1
m
# OR, to make it concise
m = matrix(0, sz, sz)
inds = rbind(cbind(1:(sz-1), 2:sz), cbind(2:sz, 1:(sz-1)))
replace(m, inds, 1)
# [,1] [,2] [,3] [,4] [,5]
#[1,] 0 1 0 0 0
#[2,] 1 0 1 0 0
#[3,] 0 1 0 1 0
#[4,] 0 0 1 0 1
#[5,] 0 0 0 1 0
We could create a function using a math trick which would work for all square matrix.
get_off_diagonal_1s <- function(n) {
#Create a matrix with all 0's
mat <- matrix(0, ncol = n, nrow = n)
#Subtract row indices by column indices
inds = row(mat) - col(mat)
#Replace values where inds is 1 or -1
mat[inds == 1 | inds == -1] = 1
mat
}
get_off_diagonal_1s(5)
# [,1] [,2] [,3] [,4] [,5]
#[1,] 0 1 0 0 0
#[2,] 1 0 1 0 0
#[3,] 0 1 0 1 0
#[4,] 0 0 1 0 1
#[5,] 0 0 0 1 0
get_off_diagonal_1s(8)
# [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8]
#[1,] 0 1 0 0 0 0 0 0
#[2,] 1 0 1 0 0 0 0 0
#[3,] 0 1 0 1 0 0 0 0
#[4,] 0 0 1 0 1 0 0 0
#[5,] 0 0 0 1 0 1 0 0
#[6,] 0 0 0 0 1 0 1 0
#[7,] 0 0 0 0 0 1 0 1
#[8,] 0 0 0 0 0 0 1 0
I have a smaller kxk matrix m given from which I want to create multiple larger NxN diagonal block matrices Q1, Q2, ..., QN. It is ensured that N is always a multiple of k.
A simple example should illustrate better what I mean:
m <- matrix(c(1,3,2,4),2,2) # the small kxk matrix
m
[,1] [,2]
[1,] 1 2
[2,] 3 4
And I want to get for let's say a 6x6 matrix the following diagonal block matrices:
Q1
[,1] [,2] [,3] [,4] [,5] [,6]
[1,] 1 2 0 0 0 0
[2,] 3 4 0 0 0 0
[3,] 0 0 0 0 0 0
[4,] 0 0 0 0 0 0
[5,] 0 0 0 0 0 0
[6,] 0 0 0 0 0 0
Q2
[,1] [,2] [,3] [,4] [,5] [,6]
[1,] 0 0 0 0 0 0
[2,] 0 0 0 0 0 0
[3,] 0 0 1 2 0 0
[4,] 0 0 3 4 0 0
[5,] 0 0 0 0 0 0
[6,] 0 0 0 0 0 0
Q3
[,1] [,2] [,3] [,4] [,5] [,6]
[1,] 0 0 0 0 0 0
[2,] 0 0 0 0 0 0
[3,] 0 0 0 0 0 0
[4,] 0 0 0 0 0 0
[5,] 0 0 0 0 1 2
[6,] 0 0 0 0 3 4
Any ideas how I could achieve this e.g. with lapply such that I can do the same for large matrices?
We can do this with bdiag from Matrix
library(Matrix)
lst <- list(bdiag(m, diag(4)*0), bdiag(0*diag(2), m, 0*diag(2)), bdiag(diag(4)*0, m))
If we want to change it to matrix, then use as.matrix
lapply(lst, as.matrix)
Also, this can be created as a single sparseMatrix
bdiag(list(m, 0*diag(6))[rep(1:2, length.out=5)])
I am working on a multi-band Raster image. For the processing Raster is converted to matrix and then pixel wise processed. I am facing problem in writing the pixel values to the blank 3d array which is explained below.
let us assume
I have a blank 3D array
x = array(0,c(4,5,3))
x
, , 1
[,1] [,2] [,3] [,4] [,5]
[1,] 0 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
, , 2
[,1] [,2] [,3] [,4] [,5]
[1,] 0 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
, , 3
[,1] [,2] [,3] [,4] [,5]
[1,] 0 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
and a vector
y = c(5,14)
I want to copy values of y in the 3rd dimension of x so that output look like example given below
x
, , 1
[,1] [,2] [,3] [,4] [,5]
[1,] 5 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
, , 2
[,1] [,2] [,3] [,4] [,5]
[1,] 14 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
, , 3
[,1] [,2] [,3] [,4] [,5]
[1,] NA 0 0 0 0
[2,] 0 0 0 0 0
[3,] 0 0 0 0 0
[4,] 0 0 0 0 0
I have searched on the internet but I am not able to find any solution.
I am trying to create a function that takes in a matrix M and a vector v. It should then take the elementwise minimum between columns of M and v. As such, the number of rows of M = length(v)
For example, the below does it for two vectors of equal length. I want it to work for a matrix compared to a vector.
vectorelementwisemin = function(x,y){ #x is a vector, y is a vector (same length)
ind = which(x > y)
z = x
z[ind] <- y[ind]
return(z)
}
For example, the vectorized function could take in:
M
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12]
[1,] 0 0 0 0 0 0 0 0 0 0 0 0
[2,] 2 2 5 4 2 3 4 1 4 4 4 2
[3,] 0 0 0 0 0 0 0 0 0 0 0 0
[4,] 0 0 0 0 0 0 0 0 0 0 0 0
[5,] 0 6 0 0 0 0 0 0 0 0 0 0
[6,] 0 0 0 0 0 0 0 0 0 0 0 0
v
0 4 2 1 3 0
And return
minmat(M,v)
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12]
[1,] 0 0 0 0 0 0 0 0 0 0 0 0
[2,] 2 2 4 4 2 3 4 1 4 4 4 2
[3,] 0 0 0 0 0 0 0 0 0 0 0 0
[4,] 0 0 0 0 0 0 0 0 0 0 0 0
[5,] 0 3 0 0 0 0 0 0 0 0 0 0
[6,] 0 0 0 0 0 0 0 0 0 0 0 0
you can just use minmat <- function(M, v) pmin(M, v) although you may want to add something like if (nrow(M) != length(v)) stop("")