Split delimited set into N-mentions+others

[jamesrkg]

Assuming data in a delimited set variable, where we know the response data is ordered, such as:

6;56;3;40;22;21;36;1;12;43;11;18;15;17;27;25;48;57;55;62;

It would be good to be able split the variable into N+1 new variables where N is the number of single-response mentions from the first, with all mentions beyond N ending up in a delimited set with all other responses.

This will let us transform ordered delimited set data into first/second/third/etc. mentions + other mentions.

[alextanski]

signature will be: nfirst(ds, name, n_first=3, others='others', reduce_values=False)

[alextanski]

Docstring not yet in, therefore parameters:

• name: the name of the delimited set.
• n_first: get the first n mentions from the code string.
• other: naming suffix for variable name & label ~(label is not good, I copy and pasted from the variable name; it should be more like '(others)', also true for the mentions...).~ If None, there won't be an other-variable.
• reduce_values: drop all metadata based on codes_in_data, i.e. get rid of empty categories.

[alextanski]

Hell, I simply fixed the variable labels quickly.

[alextanski]

Clean-up needed, code has some smaller redundancies.

[alextanski]

Docstring still needed.

[alextanski]

all done. closed.