从数据框中删除NA，但在R列中除外

Question

I have a dataframe:

column1 <- c(400, 500, 350, 70, NA, NA, 500, 350, 70)
column2 <- c(NA, 67, 23, 45, 78, 123, 54, 123, 898)
column3 <- c(412, 4, 5, 6, 1, 689, 90, 12, NA)
column4 <- c("IMC", "VMC", NA, NA, NA, NA, "IMC", "VMC", "IMC")
data1 <- data.frame(column1, column2, column3, column4)

It looks like this:

column1  column2  column3  column4
400      NA       412      IMC
500      67       4        VMC
350      23       5        NA
70       45       6        NA
NA       78       1        NA
NA       123      689      NA
500      54       90       IMC
350      123      12       VMC
70       898      NA       IMC

Question: Is it possible to remove the rows with NA's from column 1, 2 and 3, but not from column 4? So my output will look like this:

column1  column2  column3  column4
500      67       4        VMC
350      23       5        NA
70       45       6        NA
500      54       90       IMC
350      123      12       VMC

I have been struggling with this for a while now. Thanks!

Answer 1

You can use complete.cases() on the first three columns and then do a subset :

data1[complete.cases(data1[, 1:3]), ]
  column1 column2 column3 column4
2     500      67       4     VMC
3     350      23       5    <NA>
4      70      45       6    <NA>
7     500      54      90     IMC
8     350     123      12     VMC

Answer 2

A library(dplyr) approach :

data1 %>% filter(complete.cases(.[,-4]))

or

data1 %>% filter(complete.cases(.[,-3]))

if you want to keep NA's in column 3

Answer 3

Here is an option using data.table by specifying the columns in .SDcols

library(data.table)
setDT(data1)[data1[, complete.cases(.SD), .SDcols = column1:column3]]
#   column1 column2 column3 column4
#1:     500      67       4     VMC
#2:     350      23       5      NA
#3:      70      45       6      NA
#4:     500      54      90     IMC
#5:     350     123      12     VMC