Filling in NAs in data in R by id - r

Question

I am having a little problem "filling in gaps". It's not a missing data question, it's more about merging but it's not working great.
So, my data looks like this
id name region Company
1 John Smith West Walmart
1 John Smith West Amazon
1 John Smith
1 John Smith West P&G
2 Jane Smith South Apple
2 Jane Smith
3 Richard Burkett
3 Richard Burkett West Walmart
And so on.
What I want to do is fill in those gaps in the region variable by their id. So, id 1, John Smith, on the third row, should have West in the third column. Jane Smith's region should be filled in "South" where it is missing.
I've tried creating a separate dataset and then merging it based on id but it creates duplicate rows and basically increases the N by something like 14 times (no idea why).
region1<-subset(df1, df1$region=="DC"| df1$region=="Midwest"|df1$region=="Northeast"|df1$region=="South"|df1$region=="West")
region<-region1[,c(id","region")]
df2<-merge(df1, region, by="id")
I've checked the structure of the variables. Id variable is interval and region is a factor. I think there should be a super simple way to do this but I'm just not getting it. Any ideas?
Thank you in advance.

Answer 1

Here´s an R base solution. Suppose your data.frame is df
regions <- sapply(split(df$region, df$id), function(x) {
ind <- is.na(x);
x[ind] <- x[!ind][1];
x
})
df$region <- unlist(regions)
df
id name region Company
1 1 John Smith West Walmart
2 1 John Smith West Amazon
3 1 John Smith West <NA>
4 1 John Smith West P&G
5 2 Jane Smith South Apple
6 2 Jane Smith South <NA>
7 3 Richard Burkett West Walmart
8 3 Richard Burkett West <NA>

Answer 2

I would use dplyr::arrange followed by tidyr::fill
library(dplyr)
library(tidyr)
data.frame(id=c(1,1,1,1,2,2,3,3),
name=c(rep("John Smith",4), rep("Jane Smith", 2), rep("Richard Burkett", 2)),
region=c("West", "West", NA, "West", "South",NA, "West", NA),
Company=c("Walmart","Amazon",NA,"P&G","Apple",NA,"Walmart",NA)) %>%
arrange(id, name) %>%
fill(region)
Results in:
id name region Company
1 1 John Smith West Walmart
2 1 John Smith West Amazon
3 1 John Smith West NA
4 1 John Smith West P&G
5 2 Jane Smith South Apple
6 2 Jane Smith South NA
7 3 Richard Burkett West Walmart
8 3 Richard Burkett West NA

Answer 3

The solution which should work is group_by on id and then fill. Ideally the solution which should work in OP condition should cover in both direction.
library(tidyverse)
df %>% group_by(id) %>%
fill(region) %>%
fill(region, .direction = "up")
# id name region Company
# <int> <chr> <chr> <chr>
#1 1 John Smith West Walmart
#2 1 John Smith West Amazon
#3 1 John Smith West <NA>
#4 1 John Smith West P&G
#5 2 Jane Smith South Apple
#6 2 Jane Smith South <NA>
#7 3 Richard Burkett West Walmart
#8 3 Richard Burkett West <NA>
Data
structure(list(id = c(1L, 1L, 1L, 1L, 2L, 2L, 3L, 3L), name = c("John Smith",
"John Smith", "John Smith", "John Smith", "Jane Smith", "Jane Smith",
"Richard Burkett", "Richard Burkett"), region = c("West", "West",
NA, "West", "South", NA, "West", NA), Company = c("Walmart",
"Amazon", NA, "P&G", "Apple", NA, "Walmart", NA)), .Names = c("id",
"name", "region", "Company"), class = "data.frame", row.names = c(NA,
-8L))