Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

R: How to match/join 2 matrices of different dimensions (nrow/ncol)?

Tags:

r

I want to match/join two matrices, a small one with values should match in a bigger one by rownames/colnames. I only find this answer. However, I cannot match the locations as the codeline frn <- as.matrix(bigMatrix[[1]]) does not work in my case. The answers of inner, outer ... join here did not help, as I want to match/join over a lot of different columns (and not e.g. a CostumerID for X and a CustomerID for y).

As the matrices I use are a 126x104 and a 193x193 matrices. I prepared example data: 1. the bigger matrix where the smaller one should be included (the letters are in the original data set country names):

a = c("A", "B", "C", "D", "E", "F")
full_matrix = matrix(nrow = length(a), ncol=length(a))
dimnames(full_matrix) <- list(levels(as.factor(a)), levels(as.factor(a)))

full_matrix
   A  B  C  D  E  F
A NA NA NA NA NA NA
B NA NA NA NA NA NA
C NA NA NA NA NA NA
D NA NA NA NA NA NA
E NA NA NA NA NA NA
F NA NA NA NA NA NA

And the smaller matrix:

matrix = matrix(c(2, 4, 3, 1, 5, 7, 3, 1, 6), nrow=3, ncol=3)
dimnames(matrix) <- list(c("B","C","E"), c("A","B","F"))

matrix
  A B F
B 2 1 3
C 4 5 1
E 3 7 6

The result should look so:

   A  B  C  D  E  F
A NA NA NA NA NA NA
B  2  1 NA NA NA  3
C  4  5 NA NA NA  1
D NA NA NA NA NA NA
E  3  7 NA NA NA  6
F NA NA NA NA NA NA
like image 760
N.Varela Avatar asked Jan 08 '23 02:01

N.Varela


1 Answers

1) as.data.frame.table Convert the small matrix to the following long form data frame L having columns Var1, Var2 and Freq and then use matrix subscripting to assign the components:

L <- as.data.frame.table(matrix)
full_matrix[as.matrix(L[1:2])] <- L$Freq

giving:

> full_matrix
   A  B  C  D  E  F
A NA NA NA NA NA NA
B  2  1 NA NA NA  3
C  4  5 NA NA NA  1
D NA NA NA NA NA NA
E  3  7 NA NA NA  6
F NA NA NA NA NA NA

2) direct subscripting

full_matrix[rownames(matrix), colnames(matrix)] <- matrix

Note: The two solutions give identical results:

a = c("A", "B", "C", "D", "E", "F")
full_matrix = matrix(nrow = length(a), ncol=length(a))
dimnames(full_matrix) <- list(levels(as.factor(a)), levels(as.factor(a)))
matrix = matrix(c(2, 4, 3, 1, 5, 7, 3, 1, 6), nrow=3, ncol=3)
dimnames(matrix) <- list(c("B","C","E"), c("A","B","F"))

fm1 <- full_matrix
L <- as.data.frame.table(matrix)
fm1[as.matrix(L[1:2])] <- L$Freq

fm2 <- full_matrix
fm2[rownames(matrix), colnames(matrix)] <- matrix

identical(fm1, fm2)
## [1] TRUE
like image 57
G. Grothendieck Avatar answered Jan 19 '23 00:01

G. Grothendieck