parallel cbind of data frames when number of rows in data frames could differ

Question

Consider the following example-

a1<-data.frame(a=c(1,2,3),b=c(4,5,6))
a2<-data.frame(a=c(5,6),b=c(7,8))
a3<-data.frame(e=c(34,26),f=c(41,65))
a4<-data.frame(e=c(13,25,567),f=c(14,57,56))

I want to cbind a1 to a3 after dropping the last row of a1, and a2 to a4 after dropping the last row of a4 to produce

  a b  e  f
1 1 4 34 41
2 2 5 26 65

and

  a b  e  f
1 5 7 13 14
2 6 8 25 57

Map(cbind, list(a1,a2),list(a3,a4)) as has been suggested elsewhere will work if only if all data frames have the same number of rows. How do I cbind after dropping extra rows in any of the constituent data frames in the cbind?

Answer 1

We can get all the dataframe in a list using mget divide them into two halves. Use them in Map, get the minimum number of rows from both the dataframe, subset those rows and cbind.

list_dfs <- mget(paste0('a', 1:4))

Map(function(x, y) {
    rows = seq_len(min(nrow(x), nrow(y)))
    cbind(x[rows, ], y[rows, ])
    },list_dfs[1:(length(list_dfs)/2)], 
      list_dfs[(length(list_dfs)/2 + 1):length(list_dfs)])

#$a1
#  a b  e  f
#1 1 4 34 41
#2 2 5 26 65

#$a2
#  a b  e  f
#1 5 7 13 14
#2 6 8 25 57