How to replace all columns in matrix with NA based on variable in another dataframe?-CodePudding

I have a large adjacency matrix and a dataframe with a variable 'missing'. I need to replace all rows in the matrix with NA, for which 'missing' = TRUE (rows match between matrix and dataframe).

Example: I have: matrix

     [,1] [,2] [,3]
[1,]    0    0    1
[2,]    1    0    1
[3,]    0    1    0

and dataframe

  ID missing
1  1   FALSE
2  2   FALSE
3  3    TRUE

I need: matrix

     [,1] [,2] [,3]
[1,]    0    0    1
[2,]    1    0    1
[3,]   NA    NA    NA

Reproducible data:

m1.data <- c(0, 1, 0, 0, 0, 1, 1, 1, 0 )
m1 <- matrix(m1.data, nrow = 3)

df <- data.frame(ID = c(1, 2, 3),
                missing = c(FALSE, FALSE, TRUE))

Hope that someone can help! Thank you so much in advance!

CodePudding user response：

We can extract the missing column and use that as the row index in m1 and assign it to NA

m1[df$missing,] <- NA

-output

> m1
     [,1] [,2] [,3]
[1,]    0    0    1
[2,]    1    0    1
[3,]   NA   NA   NA

Or we may do

> NA^(df$missing) * m1
     [,1] [,2] [,3]
[1,]    0    0    1
[2,]    1    0    1
[3,]   NA   NA   NA

CodePudding user response：

With which we could get the Id for the missing and assing a NA to m1 it is similar to akrun's approach m1[df$missing,] <- NA

m1[which(df$missing==TRUE),] <- NA

     [,1] [,2] [,3]
[1,]    0    0    1
[2,]    1    0    1
[3,]   NA   NA   NA