Merge two data frame and replace the NA value in R

Question

I have a main table(a), containing column: id, age, and sex. eg.

a <- data.frame(id=letters[1:4], age=c(18,NA,9,NA), sex=c(\"M\",\"F\",\"F\",\"M\"))
  id          


        

Answer 1

We can use data.table

library(data.table)
setDT(a)[b, agei := i.age, on='id'][is.na(age), age := agei][,agei:= NULL][]
a
 #  id age sex
#1:  a  18   M
#2:  b  32   F
#3:  c   9   F
#4:  d  20   M

Answer 2

Here is a dplyr solution:

library(dplyr)

c <- left_join(a,b, by = "id") %>% # this will generate age.x and age.y
  mutate(age = ifelse(is.na(age.x), age.y, age.x)) %>% # we generate a joint 'age' variable
  select(-age.y, -age.x) # drop the superfluous columns

> c
  id sex age
1  a   M  18
2  b   F  32
3  c   F   9
4  d   M  20

Note that this will throw you a warning that you try to join on factor levels. This is because the example data in the reproducible example was generated with stringsAsFactors = T.