R - Delete column if last value is NA

Question

I want to delete column(s) in R data frame if their last value is NA.

Example data frame is below

A   B   C   D
11  10  19  20
22  20  29  40
33  30  39  60
44  NA  NA  80

I would like to get the following output (Column B and C dropped because they had NA values in the last row)

Tried but couldn't get any solution to work. Appreciate your help.

http://stackoverflow.com/questions/12454487/remove-columns-from-dataframe-where-some-of-values-are-na — Alexey Ferapontov, Nov 11 '16 at 20:31

JasonWang · Accepted Answer · 2016-11-11T20:49:34.137

3

# Data
df <- read.table(text="A   B   C   D
                       11  10  19  20
                       22  20  29  40
                       33  30  39  60
                       44  NA  NA  80", header=TRUE)

df[-which(is.na(df[nrow(df), ]))]

   A  D
1 11 20
2 22 40
3 33 60
4 44 80

# Suggestion from @alistaire
df[, !is.na(df[nrow(df), ])]

edited Nov 11 '16 at 20:49

answered Nov 11 '16 at 20:34

JasonWang

2,414
11
12

You can just use `!` instead of `-which`. – alistaire Nov 11 '16 at 20:37
3

Alternately, `tail(df, 1)` like `df[, !is.na(tail(df, 1)) ]` – Frank Nov 11 '16 at 20:37
1

Sorry for the mistake. It should be `nrow(df)`. – JasonWang Nov 11 '16 at 20:39
Thanks Jason, Frank and alistaire – manoj Nov 11 '16 at 20:48

score 0 · Answer 2 · answered Nov 12 '16 at 06:49

0

na.omit() function deletes any rows with missing data. pp <- data.frame(po=c(10,20,30,40,NA), pi=c(45,65,75,55,85)) pp na.omit(pp) -- deletes the entire row

na.delete(pp["po"]) --- deletes the entire column with "NA".

answered Nov 12 '16 at 06:49

Mahendra

19
2

This is true, but the OP needs to test if *the last row* is `NA` and only delete the column in that case. This answer will delete if the `NA` occurs anywhere. – Gregor Thomas Nov 12 '16 at 07:37

R - Delete column if last value is NA

2 Answers2