lapply 或 sapply 用于列表中的 data.frames

Question

I play around with Alternatives for the dplyr way of summarising data. I like the split and apply approach but need some help.

library(Hmisc)
library(data.table)

summary <- function(x) {
    funs <- c(wtd.mean, wtd.var)
    sapply(funs, function(f) f(x, na.rm = TRUE))
}


df <- split(mtcars, f = mtcars$cyl)

store <- list()

for(i in 1:length(df)) {
    store[[i]] <- data.frame(sapply(df[[i]], summary)) 
}

finaldf <- data.table::rbindlist(store)

finaldf

Here is my code. With the split function i get three dataframes with summarised values. But after that my code gets a little bit messy with creating an empty list, converting the matrix to a data.frame inside the loop etc.

Is there a way to use multiple apply functions to avoid this loop? Something like lapply(sapply(...)) ?

Answer 1

We can use lapply and avoid the initialization of list

library(data.table)
lst <- lapply(df,  function(dat) data.frame(lapply(dat, summary)))
rbindlist(lst, idcol = 'grp')
#   grp       mpg cyl      disp         hp      drat        wt      qsec         vs        am      gear      carb
#1:   4 26.663636   4  105.1364   82.63636 4.0709091 2.2857273 19.137273 0.90909091 0.7272727 4.0909091 1.5454545
#2:   4 20.338545   0  722.0825  438.25455 0.1335691 0.3244028  2.830622 0.09090909 0.2181818 0.2909091 0.2727273
#3:   6 19.742857   6  183.3143  122.28571 3.5857143 3.1171429 17.977143 0.57142857 0.4285714 3.8571429 3.4285714
#4:   6  2.112857   0 1727.4381  588.57143 0.2266286 0.1269821  2.913390 0.28571429 0.2857143 0.4761905 3.2857143
#5:   8 15.100000   8  353.1000  209.21429 3.2292857 3.9992143 16.772143 0.00000000 0.1428571 3.2857143 3.5000000
#6:   8  6.553846   0 4592.9523 2598.64286 0.1386533 0.5766956  1.430449 0.00000000 0.1318681 0.5274725 2.4230769

---

The steps can be much simplified as well if we use data.table group by methods

as.data.table(mtcars)[, lapply(.SD, summary), by = cyl]

Or instead of sapply ing the functions, apply it individually and concatenate the output

summary1 <- function(x)  c(wtd.mean(x, na.rm = TRUE), wtd.var(x, na.rm = TRUE))
as.data.table(mtcars)[, lapply(.SD, summary1), by = cyl]