How can I write one line code instead of taking mean of each variable inside the group_by()?

My code is following:

mtmm_data4 <- mtmm_data  %>% group_by(cntry_lan, admdw) %>% 
  summarise( M1T1 = mean(M1T1, na.rm = TRUE), M1T2 = mean(M1T2, na.rm = TRUE), M1T3 = mean(M1T3, na.rm= TRUE), M2T1 = mean(M2T1, na.rm= TRUE), M2T2 = mean(M2T2, na.rm=TRUE), M2T3 = mean(M2T3, na.rm=TRUE), M3T1 = mean(M3T1, na.rm), M3T2 = mean(M3T2, na.rm=TRUE), M3T3 = mean(M3T3, na.rm = TRUE))

Also this code created a data set where all M2T1, M3T1, M3T2 are NANs. I did not understand why I would have had such NAs.

>Solution :

 mtmm_data  %>% 
   group_by(cntry_lan, admdw) %>% 
   summarise(across(c(YOUR_COLUMNS), ~mean(., na.rm = TRUE)))

Have a look at the tidyselect functions for some helpers that might make it simpler / more flexible to select the columns you want to summarize.

EDIT —

It sounds like your data might have been read in as character instead of numeric, which would error out when you take the mean. Examine your data types (e.g. str(mtmm_data) or glimpse(mtmm_data)) and if that’s the case, the best thing would be to fix it upstream when you load the data. You could also do it here with something like:

mtmm_data  %>% 
   group_by(cntry_lan, admdw) %>% 
   summarise(across(c(YOUR_COLUMNS), ~mean(as.numeric(.), na.rm = TRUE)))