Using summarise_at(). Weighted mean Tidyverse approach

Supose you are analysing survey data. You are asked to get the mean in a representative way, weighting your individuals depending on the number of members of their segment.

library(tidyverse)

survey_data <- tribble(
  ~id, ~region1, ~region2, ~gender, ~q1, ~q2,
  1,"sp","mad","m", 2,5,
  2,"it", "bol", "m", 5, 10,
  3,"sp", "bar", "f", 2, 2,
  4,"sp", "bar", "f", 7, 7,
  5,"it", "bol", "m", 2, 7)  

  survey_data %>% 
  group_by(region1, region2, gender) %>% 
  mutate(weight = 1/n()) %>% 
  ungroup() %>% 
  summarise_at(vars(contains("q")),
               funs(weighted_mean = sum(. * weight)/sum(weight)))