How can I mutate data based on whether a string in R contains a specific value?

For merging purposes, I summarized rows of an existing dataset (structured by country, year and party) as follows, as I am not interested in the values of distinct parties, but whether a distinct kind of parties were part of the government in a year:

library(dplyr)

vparty <- vparty%>%
  group_by(country_name, year) %>%
  summarise(across(everything(), ~toString(.)))

Now I have a data frame looking like this:

What I want to do now is, to mutate apingov depending on whether the string contains a "1" or a "0". (It is not possible that there is 1 and 0 in the same string, and the NAs are not important when 1 or 0 is contained)- For a string containing 1, the value of the variable shall be 1, for a string containing 0 = 0, when there is neither 1 nor 0 than the value shall be NA.

I tried different solutions I found here. However, none of them really worked for my specific case.

You can use parse_number from the readr package for that. I have added an extra row demonstrating the behaviour when all apingov is NA.

library(dplyr)

vparty %>% mutate(apingov = parse_number(apingov))

If there are numbers other than "1" or "0", the above parse_number function would not report NA. Use grepl in case_when if that's the case.

vparty %>% 
  mutate(apingov = case_when(grepl("1", apingov) ~ "1",
                             grepl("0", apingov) ~ "0",
                             .default = NA))

Output

  country_name  year apingov
1     Country1 year1       1
2     Country1 year2       0
3     Country2 year1       1
4     Country2 year2       0
5     Country2 year2      NA

Data

vparty <- structure(list(country_name = c("Country1", "Country1", "Country2", 
"Country2", "Country2"), year = c("year1", "year2", "year1", 
"year2", "year2"), apingov = c("NA, NA, NA, 1, NA", "NA, NA, NA, 0, NA", 
"NA, 1, NA, NA, NA", "NA, NA, NA, NA, 0", "NA, NA, NA, NA, NA"
)), class = "data.frame", row.names = c(NA, -5L))

I would recommend parse_number. Here is a version with str_detect and case_when

library(dplyr)
library(stringr)

vparty %>% 
  mutate(apingov = case_when(
    str_detect(apingov, "1") ~ 1,
    str_detect(apingov, "0") ~ 0,
    TRUE ~ NA_real_
  ))

 country_name  year apingov
1     Country1 year1       1
2     Country1 year2       0
3     Country2 year1       1
4     Country2 year2       0