1

I am trying to use the recode into function using expss. When I replicate the example illustrated in https://gdemin.github.io/expss/#example_of_data_processing_with_multiple-response_variables I notice the following behaviour: code including a do_if statement works perfect:

w = w %>%  do_if(cell == 1, 
    {recode(a1_1 %to% a1_6, other ~ copy) %into% (h1_1 %to% h1_6)})

Without the do_if statement

w = w %>% recode(a1_1 %to% a1_6, other ~ copy) %into% (h1_1 %to% h1_6)

an error is generated "Error: 'a1_1' not found".

Is this correct?

It would also be nice to transfer variabele labels to the newly created variables (cfr. mutate)

Thanks for any advice on how to use correctly.

michaela stubbers
  • 61
  • 3

1 Answers1

0

Inside do_if recode is evaluated in the context of the data.frame. When there is no context recode will be evaluated in the global environment. And it don't see a1_1,... variables in the data.frame. The good news is that there is a compute function for arbitrary operations on data.frames:

w = w %>%  
    compute({
         recode(a1_1 %to% a1_6, other ~ copy) %into% (h1_1 %to% h1_6)
    })

As for labels copying, you can use argument 'with_labels' = TRUE or an alias 'rec' which has 'with_labels' = TRUE by default:

w = w %>%  
    compute({
         rec(a1_1 %to% a1_6, other ~ copy) %into% (h1_1 %to% h1_6)
    })
Gregory Demin
  • 4,596
  • 2
  • 20
  • 20