Remove lines with only NAs from data.table

Question

I want to remove the lines from a data.table that only contain NAs.

> tab = data.table(A = c(1, NA, 3), B = c(NA, NA, 3))
> tab
    A  B
1:  1 NA
2: NA         


        

Answer 1

I quite like

tab <- tab[sapply(1:nrow(tab), function(i){!all(is.na(tab[i,]))}),]

It is intuitive to me, but I'm not sure it is the fastest approach.

HTH

Answer 2

We can use Reduce with is.na and &

tab[!Reduce(`&`, lapply(tab, is.na))]
#   A  B
#1: 1 NA
#2: 3  3

---

Or a compact but not so efficient approach would be

tab[rowSums(!is.na(tab)) != 0L]

Also, as commented by @Frank, a join based approach,

tab[!tab[NA_integer_], on = names(tab)]

Answer 3

Another idea:

library(dplyr)
df %>% 
  filter(rowSums(is.na(.)) < length(.))