1

I want to create a weighted mean of a dichotomous variables (0 or 1) using expss packages. AND the results will be presented in percentage. These dichotomous variables contain missing values.

fisrt of all, i create a function :

tableau = function (Q, banner,titre) {

banner <- df %>% 
  tab_cols(a2,a3)

    eval.parent(substitute(
  {
    banner %>%
      tab_cells (list(Q)) %>%
   tab_weight(weighting_var) %>%
    tab_stat_fun("%" = function (x)  mean(x,na.rm = TRUE)*100) %>%
     tab_pivot() %>% 
      drop_empty_columns() %>%
      tab_transpose() %>% 
      set_caption(titre) #%>%
    # custom_format() 
  }
))
}

When i use this function:

tableau(list(i1.3_1,i1.3_2,i1.3_3,i1.3_4,i1.3_5,i1.3_6),banner,'TABLE 1')

i get this message :

`weight` is provided but function`%` doesn't have formal `weight` argument

how can I modify my function to make sure that the results are weighting?

desertnaut
  • 57,590
  • 26
  • 140
  • 166
Legarraudien
  • 21
  • 3

1 Answers1

0

You need formal weight parameter in your function. In your case it looks like this: "%" = function (x, weight) w_mean(x, weight = weight, na.rm = TRUE)*100

Gregory Demin
  • 4,596
  • 2
  • 20
  • 20