R: how to get row and column names of the true elements of a matrix?

Question

I have a logical matrix x with named rows (\'a\' and \'b\') and named columns (\'10\', \'20\', \'30\', \'40\'). Let\'s say, this:

   10   20   3         


        

Answer 1

You can use the fact that table objects are converted to "long" format data frames by as.data.frame():

# Create matrix of interest 
mat <- structure(c(TRUE, FALSE, FALSE, TRUE, TRUE, FALSE, FALSE, TRUE), .Dim = c(2L, 4L), .Dimnames = list(c("a", "b"), c("10", "20", "30", "40")))

# Convert to table, then to long data.frame
df <- mat %>% as.table %>% as.data.frame(., stringsAsFactors=FALSE)

The resulting df is the following:

  Var1 Var2  Freq
1    a   10  TRUE
2    b   10 FALSE
3    a   20 FALSE
4    b   20  TRUE
5    a   30  TRUE
6    b   30 FALSE
7    a   40 FALSE
8    b   40  TRUE

Which you can then index to only keep TRUE rows:

df <- df[df$Freq,1:2] %>% sort
df
      Var1 Var2
    1    a   10
    5    a   30
    4    b   20
    8    b   40

You can use dplyr to convert this into exactly the table you want:

library(plyr)
ddply(df, "Var1", function(x) x$Var2)
  Var1 V1 V2
1    a 10 30
2    b 20 40