Answer to: Calculate multilple stats for two or more columns in R data.table

You need to modify stats to accept an array (one column per numeric variable) and return a simple list (not a list of lists). Keeping everything as lists (instead of arrays or matrices) within stats will preserve the names needed to identify the columns. stats <- function(x) { stat <- function(x) { y <- quantile(x, c(1/4, 1/2, 3/4)) names(y) <- c("Q1", "M", "Q3") as.list(y) } if (!is.matrix(x)) x <- matrix(x, ncol = 1) as.list(unlist(apply(x, 2, stat))) } When using more than one value column, create the array argument to stats with cbind: dt[, c(stats(cbind(val1, val2))), keyby = .(class)] class val1.Q1 val1.M val1.Q3 val2.Q1 val2.M val2.Q3 <char> <num> <num> <num> <num> <num> <num> 1: a 8.0 9.5 10.25 67.5 95 112.5 2: b 2.5 3.5 4.75 55.0 65 77.5 3: c 5.0 7.0 9.00 17.5 35 60.0