Count valid/non-NA observations grouped by variable

I have been googling for about two hours now trying to find the solution to a simple problem, but I am not getting anywhere. I am not even finding the right function I could use to get where I want to get, so I have to resort to asking for help even though it seems a very basic question.

I have a survey spanning various countries, and I am in the process of creating a dataframe with statistics derived from the data (such as national mean on a given variable).

Let's say the dataframe df looks like this:

country <- c(1,1,1,1,1,1,2,2,2,2,2,2)
var <- c(1,NA,2,1,2,2,3,3,1,3,4,NA)
df <- cbind.data.frame(country, var)

I can easily count the nominal n's:

df %>% group_by(country) %>% summarize(n=n())

But how do I count valid observations on the variable var?

Any of these will do it.
Three function calls, group_by()/summarise()/n() can be replaced by one only, count().

country <- c(1,1,1,1,1,1,2,2,2,2,2,2)
var <- c(1,NA,2,1,2,2,3,3,1,3,4,NA)
df <- cbind.data.frame(country, var)
suppressPackageStartupMessages(
  library(dplyr)
)
df %>% 
  na.exclude() %>%
  count(country)
#>   country n
#> 1       1 5
#> 2       2 5
df %>% 
  na.omit() %>%
  count(country) 
#>   country n
#> 1       1 5
#> 2       2 5
df %>% 
  tidyr::drop_na() %>%
  count(country)
#>   country n
#> 1       1 5
#> 2       2 5

^{Created on 2023-08-08 with reprex v2.0.2}