Search code examples
rconditional-statementsmultiple-columnsnaaccount

the number of NAs per conditions

I have a long database df. and there are two columns city (10000 with NAs) and country (i have 250 countries). I want to know the number of NAs cities per country. I tried to do pivot wider and then colSums(is.na(df)) but is not working. Is there an other way?

many thanks Eleni

Solution

  • Without a reproducable dataset it is hard to say but using dpyr/tidyverse I'd do something like this:

    df %>%
   group_by(country) %>%
   summarise(count_na = sum(is.na(city)))