[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: graph editor stata10
No, it wasn't a joke. But it was a direct response inspired by Bill
Gould's earlier post that stata thought about limiting the editor. Bill
We looked long and hard at the issue and came to the conclusion that
restricting the editor to doing just things the editor could "understand"
(and so record generically) would overly limit the kind of things
interactive users could do.
So, I reasoned that now that the full editor has been produced,
constructing a less powerful version of the editor would be relatively
simple and would not lead users to be frustrated. I can imagine that if
they had released the limited one first, lots of complaints would have
ensued. However, if stata now makes and releases the limited one, any
complaints as to its limits can be easily answered--just use the full
I winked because it is a bit early to be making such a list. On the other
hand, something should be mentioned for the wishlist when it arises, else
it may get lost in the shuffle (and thus not really evaluated and,
perhaps, explicitly rejected). If I recall correctly, this was what
happened with stata9, as several users pointed out that providing mixed
models only for interval level variables was a weakness, and that gllamm
was not sufficient (too powerful, not fully integrated, not fully
supported) as the only means of estimating mixed models for categorical
data. Thankfully, it appears that comment need no longer be made.
On Wed, 27 Jun 2007, Nick Cox wrote:
> Somehow, I don't think this horse will run. Nor
> would I suggest that StataCorp even think about
> Stata 10 is officially two days old, although
> shipping started last week. I doubt that anyone
> outside StataCorp has as yet much experience with
> the graph editor. So, suggesting, already, that we need _another_
> one is a trifle premature, unless this is a joke.
> What I suggest is different. Anyone who repeatedly
> finds themselves making the same kinds of changes
> to graphs in the Graph Editor can make that clear.
> We will then find that
> 1. There is already a way of automating the production
> of graphs with those changes. (You've not registered
> that an option exists to do your tweak.) Or, alternatively, there
> is another way of producing the graph you want.
> (You are using -graph bar-, and you would be better
> off with -twoway bar-, or whatever.)
> 2. There isn't a way, but StataCorp will find that they
> can tweak the graph language upstream to make that true.
> 3. There isn't, and StataCorp won't or can't do
> anything about it, usually because the changes really aren't
> automatable in any reasonable way.
> My very wild guess is that the distribution
> will be #1 80%; #2 10%; #3 10%.
> This does not refer to anything that people
> do to graphs outside Stata, whether it is to do with
> Greek symbols, polka dots or drawings of
> little people.
> > Sounds like a nice addition for Stata11 (eleven) would be PGHE-a
> > Programmer's Graph Helper Editor--which would be the same as
> > the Stata10
> > editor *except* it would be restricted to the things that a programmer
> > could use syntax to write. Then, a programmer could program
> > a graph in a
> > do-file, edit it using PGHE, and then take the new syntax and
> > run as many
> > versions as s/he wanted. And, the existence of the existing
> > Stata10 graph
> > editor (which would be retained in Stata11) would allow much
> > more flexible
> > editing.
> > So, that's wishlist item #2 for Stata11. ;-)
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
* For searches and help try: