[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: an ordinary two-way table
This sounds like a nice Stataish way of solving this
particular problem. Any value you care enough about to define
a value label for (even of a limiting form 1 "1" 2 "2"
etc.) is programmably accessible by Stata.
The general issue remains, floating somewhere between
platitude and profundity....
> I agree with much of what you say, as usual, but it seems to me that
> -tabulate- could have an option -label- (much like option -missing-)
> that would make sure that every category of a numeric variable that
> has a value label defined for it would be included in the table. Then
> many of these odd tables could be constructed in 2 or 3 easy steps. I
> can envision an easy ado-based solution not using Mata, but using
> -label save- and -file read- loops to read off the 4th "word" of the
> do-file that results from -label save-. If I have a chance, I'll
> write up a quickie version later today...
> You could even make a version that makes SAS-like tables, where
> categories with the same label are collapsed into the same row or
> column, but I doubt I would provide that functionality (call me
> On 5/11/07, Nick Cox <firstname.lastname@example.org> wrote:
> > The problem with tables is analogous to that of
> > graphs.
> > In essence tabulation in Stata is a bit of a mess
> > but then what is the alternative? One enormous command
> > with numerous bells and whistles and
> > with a manual entry 150 pages long that will do
> > all these special things too?
* For searches and help try: