Fill NA in R: imput NA in a column X with values from same ID (column Y ) correspondance

Question

I´ve got this dataset (something like below)

db <- as.data.frame(cbind(c(1,2,2,3,3,4,4,4,5,5),c('a','b',NA,NA,'i',NA,'d',NA, NA, NA)))

I´d like to fill the V2 NA with same correpondance from ID column V1. At the end I expect this result

db <- as.data.frame(cbind(c(1,2,2,3,3,4,4,4,5,5),c('a','b','b','i','i','d','d','d', NA, NA)))

I´ve tried to make a list with unique ID db_aux <- as.data.frame(cbind(c(1,2,3,4),c('a','b','i','d')))

but I guess it´s necessary an apply function to fill in what´s left but i´m not figuring how to indicate the corresponding indexation.

Answer 1

You can use fill from tidyr with the direction being updown or downup.

library(tidyverse)

db %>%
   group_by(V1) %>%
   fill(V2, .direction = 'updown')

# A tibble: 10 x 2
# Groups:   V1 [5]
   V1    V2   
   <chr> <chr>
 1 1     a    
 2 2     b    
 3 2     b    
 4 3     i    
 5 3     i    
 6 4     d    
 7 4     d    
 8 4     d    
 9 5     NA   
10 5     NA