Deal with occasional numeric(0) inside group

I have a dataset of ids, months and some metric x.

library(tidyverse)
library(lubridate) 

df <- data.frame(
  id = c("a1", "a1", "a1", "b1", "b1", "b1"), 
  month = c("2021-01-01", "2021-03-01", "2021-04-01", "2021-05-01", "2021-06-01", "2021-08-01"), 
  x = c(34,56,76, 12, 13, 14), 
  month_start = c("2021-01-01", "2021-01-01", "2021-01-01", "2021-05-01", "2021-05-01", "2021-05-01")
)

df <- df %>% mutate(month = as.Date(month), month_start = as.Date(month_start))

and would like to create new columns that are values of the 2nd and 4th months' x of each customer. I tried the code below but it failed because for id == "a1", the 2nd month's x is NA. (error: Input x_growth_start can't be recycled to size 3.)

df %>%
  group_by(id) %>%
  mutate(x_growth_start = x[month == month_start %m % months(1)],
         x_growth_end = x[month == month_start %m % months(3)])

I realize what it actually returns when the value doesn't exist is integer(0). Can I make the code run (say, by letting it return NA if integer(0))? I tried what was suggested here but it didn't work Input `typ` can't be recycled to size in R

Thanks in advance.

CodePudding user response：

You should use match(). It returns NA when no match is found.

df %>%
  group_by(id) %>%
  mutate(x_growth_start = x[match(month_start %m % months(1), month)],
         x_growth_end   = x[match(month_start %m % months(3), month)])

#   id    month          x month_start x_growth_start x_growth_end
#   <chr> <date>     <dbl> <date>               <dbl>        <dbl>
# 1 a1    2021-01-01    34 2021-01-01              NA           76
# 2 a1    2021-03-01    56 2021-01-01              NA           76
# 3 a1    2021-04-01    76 2021-01-01              NA           76
# 4 b1    2021-05-01    12 2021-05-01              13           14
# 5 b1    2021-06-01    13 2021-05-01              13           14
# 6 b1    2021-08-01    14 2021-05-01              13           14