为完整案例设置阈值以从 R 中的多个列中删除 NA

Question

There might be an easy answer to this, but I am not able to make it work. I have a data table that looks like this:

df <- data.table(t = c(1, 2, 3), a = c(NA, NA, 4), b = c(NA, 4, NA), c = c(NA, 4, NA))

How can I remove only the rows where all columns but "t" have NA's. It should be fast because of my big data files, so I would like to do it especially with complete.cases. I couldn't find a solution to this problem yet.

The result should look like this

dfRes <- data.table(t = c(2, 3), a = c(NA, 4), b = c(4, NA), c = c(4, NA))

Answer 1

We can use rowSums on columns other than "t" .

library(data.table)

cols <- which(names(df) != 't')
df[rowSums(!is.na(df[, ..cols])) > 0, ]

#   t  a  b  c
#1: 2 NA  4  4
#2: 3  4 NA NA

Answer 2

We can use complete.cases with Reduce

library(data.table)
df[df[, Reduce(`|`, lapply(.SD, complete.cases)), .SDcols = a:c]]
#   t  a  b  c
#1: 2 NA  4  4
#2: 3  4 NA NA