Convert columns of arbitrary class to the class of matching columns in another data.table

This is one very crude way to ensure common classes:

library(magrittr)

cols = intersect(names(A), names(B))
r    = rbindlist(list(A = A, B = B[, ..cols]), idcol = TRUE)
r[, (cols) := lapply(.SD, . %>% as.character %>% type.convert), .SDcols=cols]
B[, (cols) := r[.id=="B", ..cols]]
A[, (cols) := r[.id=="A", ..cols]]

sapply(A, class); sapply(B, class)
#      year   stratum 
# "integer" "integer" 
#      year   stratum        yr 
# "integer" "integer" "numeric" 

I don't like this solution:

• I routinely use all-integer codes for IDs (like "00001", "02995"), and this would coerce those to actual integers, which is bad.
• Who knows what this will do to fancy classes like Date or factor? This won't matter so much if you do this col-classes normalization as soon as you read data in, I suppose.

---

Data:

# slightly tweaked from OP
A <- setDT(structure(list(year = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), stratum = c(1L, 2L, 
3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 13L, 14L, 15L, 1L, 2L, 3L, 4L, 5L, 
6L, 7L, 8L, 9L, 10L, 11L, 12L, 13L, 14L, 15L, 1L, 2L, 3L, 4L, 5L, 6L, 7L, 8L, 
9L, 10L, 11L, 12L, 13L, 14L, 15L)), .Names = c("year", "stratum"), row.names = 
c(NA, -45L), class = c("data.frame")))

B <- setDT(structure(list(year = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 
1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 
3, 3, 3, 3), stratum = c(1L, 2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 13L, 
14L, 15L, 1L, 2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 13L, 14L, 15L, 1L, 
2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 13L, 14L, 15L), yr = c(1, 1, 1, 
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 
3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3)), .Names = c("year", "stratum", 
"yr"), row.names = c(NA, -45L), class = c("data.frame")))

Comment. If you have something against magrittr, use function(x) type.convert(as.character(x)) in place of the . %>% bit.

Not very elegant but you may 'build' the as.* call like this:

for (x in colnames(A)) { A[,x] <- eval( call( paste0("as.", class(B[,x])), A[,x]) )}