Concatenating all columns for each row as a string, and writing it into another data frame in R

Question

I am trying to change multiple elements of a data frame, concatenate all columns for each row and write them into a new data frame in R. To change multiple elements using dictionary, I followed Ramnath's solution in here.

You can see my sample code below:

arr <- c(1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 1073741824)

dict = list('1' = 'germline', '2' = 'somatic', '4' = 'inherited', '8' = 'paternal',
            '16' = 'maternal', '32' = 'de-novo', '64' = 'biparental', '128' = 'uniparental',
            '256' = 'not-tested', '512' = 'tested-inconclusive', '1024' = 'not-reported',
            '1073741824' = 'other')

a <- data.frame(t(combn(arr, 3)), stringsAsFactors = FALSE)
nr = nrow(a)
b <- as.data.frame(matrix(nrow = nr, ncol = 1))
row.names(b) <- rowSums(a)
a[] <- lapply(a, as.character)

for (j in 1:length(dict)) {
  a <- replace(a, a == names(dict[j]), dict[j])
}
for (i in 1:nr) {
  b[i, 1] <- paste(as.character(as.vector(a[i,])), collapse = ", ")
}

My expected output is (example):

> b[1,1]
[1] germline, somatic, inherited

However, I get this:

> b[1,1]
[1] "list(\"germline\"), list(\"somatic\"), list(\"inherited\")"

I couldn't figure out what problem is, and I would be very pleased if you could help me with it.

score 2 · Accepted Answer · answered Jan 18 '21 at 16:45

Be aware of that a[i,] is a data.frame, which can be seen from, for example

> str(a[1,])
'data.frame':   1 obs. of  3 variables:
 $ X1:List of 1
  ..$ : chr "germline"
 $ X2:List of 1
  ..$ : chr "somatic"
 $ X3:List of 1
  ..$ : chr "inherited"

One workaround is using unlist(a[i,])

for (i in 1:nr) {
  b[i, 1] <- paste(unlist(a[i, ]), collapse = ", ")
}

such that

> head(b)
                                V1
7     germline, somatic, inherited
11     germline, somatic, paternal
19     germline, somatic, maternal
35      germline, somatic, de-novo
67   germline, somatic, biparental
131 germline, somatic, uniparental

A simpler option is using do.call + paste

b[, 1] <- do.call(paste, c(a, sep = ", "))

which gives

> head(b)
                                V1
7     germline, somatic, inherited
11     germline, somatic, paternal
19     germline, somatic, maternal
35      germline, somatic, de-novo
67   germline, somatic, biparental
131 germline, somatic, uniparental

Concatenating all columns for each row as a string, and writing it into another data frame in R

1 Answers1