How to write the same value to multiple entries of an ID in a data frame in R?

My data essentially looks like this, in a much shortened form:

df <- data.frame(id = c(1,1,1,2,2,2,2,2,3,3,3,3), 
                 height = c(150, NA, 151, NA, NA, 176, 175, 174, 198, NA, 197, 198))

What I would like to do is compute the mean height for each of these IDs and then plug that height in for every NA for that given ID. So ID 1 should have a mean height of 150.5, thus the first NA should be replaced by 150.5. Then ID 2 has a mean height of 175, so I’d like to plug in 175 for the two NAs associated with ID 2. And so on.

I know I could manually enter these with things like df[2,2] <- 150.5, but in reality I have thousands of IDs and this wouldn’t be feasible.

I’m pretty comfortable with the dplyr package and I figure I should utilize group_by(id) somehow, but I can’t figure out the rest.

Any suggestions?

>Solution :

An approach using replace_na

library(dplyr)
library(tidyr)

df %>% 
  mutate(height_new = replace_na(height, mean(height, na.rm=T)), .by = id)
   id height height_new
1   1    150   150.0000
2   1     NA   150.5000
3   1    151   151.0000
4   2     NA   175.0000
5   2     NA   175.0000
6   2    176   176.0000
7   2    175   175.0000
8   2    174   174.0000
9   3    198   198.0000
10  3     NA   197.6667
11  3    197   197.0000
12  3    198   198.0000