Wishlist for Stata 15

[Sergiy Radyakin]

Really simple wish:

Can we have range (difference between max and min) reported by summarize ?
The monitors are now wide enough to accommodate a few more (imho, useful) statistics:

. summarize

    Variable |       Obs        Mean    Std. Dev.       Min        Max     Range
-------------+-------------------------------------------------------------------
           y |        14    38.89876    .0002532   38.89841    38.8993    .000887
           x |        14   -77.04115    .0002442   -77.0416  -77.04089    .000712

There are a few of stats in tabstat, that can become options for summarize as well. Some of them (like range) don't require much additional computations (range, cv) or are available in r() anyway (sum).

Would it also be possible to have tabstat *, stats(*) ?
Often faster then listing stats one after another.

Thank you, Sergiy Radyakin

[Guest]

Thank you, Sergiy Radyakin, for establishing a "Wishlist for Stata" in the General section. Maybe some wishes may be solved in future updates of version 14 instead of having to wait until version 15.

I'd like to add a few points of my own:

Very simple, but also important:

-putexcel- to include output with three decimals (i.e. nformat("number_sep_d3") in addition to the current nformat("number_sep_d2").
Maybe in Stata 14.01?

More ambitious:

- substantial speed improvement in -sem- and -gsem-. For me, sem and gsem cannot be used in the current version of Stata, due to being extremely slow.
- Graphics in -sem- as output. (Currently, one can use graphical input, which helps learning SEM, but is also a slow way of input. For the many who use code: A graphical output will sometimes help inspect more complex models developed.
- Bayes to include sem/gsem, including multilevel SEM.

Apart from that? Continue developing this wonderful software... Thanks!

Guest

[Robert Picard]

I vote for support of Stata comments (i..e "//", "///", "/* */") in the Command window. It's a pain for those who do not use the built-in editor to cut and paste code from a do-file to the Command window. The issue also applies to Statalist posts that include such comments.

[daniel klein]

I vote for support of Stata comments (i..e "//", "///", "/* */") in the Command window. It's a pain for those who do not use the built-in editor to cut and paste code from a do-file to the Command window. The issue also applies to Statalist posts that include such comments.

If that was possible, it would really be a great addition. So, plus one for this suggestion.

[daniel klein]

Could the doedit command throw back an error 601 or perhaps just a note if a specified filename was not found, and not open up a new empty do-file? I cannot think of any situation where I would be happy to see a new window popping up when I actually wanted to open an existing file. It is just inconvenient to have to close down the new window every time I have a typo in the filename or something.

Thanks.
Daniel

[Nick Cox]

Daniel:

You could program this for yourself with a mydoedit that fails when no file was found with such a name.

[daniel klein]

True, but I guess just because something is simple to program myself does not mean it should be discarded from the wishlist, should it? Sergiy once made the suggestion that this could be a good idea:

I tend to think of a wish in this context as something that is not doable by the user in principle, but something that should be relatively easy to do for developers having access to internals.

However, I do not believe everyone on the list can tell the difference between my suggestion for doedit and Robert's suggestion to allow comments in the command line. And, for that matter, I do not believe that everyone on the list can program this. So I leave it to StataCorp to decide whether to implement only those wishes that cannot be programmed by users.

Best
Daniel

[Richard Williams]

True, but I guess just because something is simple to program myself does not mean it should be discarded from the wishlist, should it?

I don't think Nick is necessarily degrading it, I think he is just saying that until Stata gets around to it here is what you can do in the meantime. One nice thing that has emerged from previous wish lists is such suggestions -- indeed sometimes you find that your wish has already been fulfilled and you didn't know it!

[Nick Cox]

Richard is right. I wasn't commenting on whether it is a good idea, but just underlining that it is programmable by users.

Here is a zeroth draft

program mydoedit
    version 14
    if "`1'" != "" {
        confirm file `1'
    }
    doedit `1'
end

I made various suggestions in the 1990s that StataCorp have not (yet) implemented, so sometimes patience is needed.

Daniel is also right. It's pretty hard even for user-programmers to spot what is feasible. I can think of a couple of projects in which I was absolutely convinced that a change to Stata's internals was needed first, until I realised how to do it otherwise.

[Alfonso Sánchez-Peñalver]

I wish....

Multinomial mixed models (probit and logit) without having to learn SEM/GSEM. Currently possible thanks to mixlogit (Hole (2007) SSC) and cmp (Roodman (2011) SSC).

References:
Hole, A.R. (2007), "Fitting Mixed Logit Models by Using Maximum Simulated Likelihood," The Stata Journal 7(3), pp. 388-401. (st0133)

Roodman, D. (2011), "Fitting Fully Observed Recursive Mixed-process models with cmp," The Stata Journal 11(2), pp. 159--206. (st0224)

[daniel klein]

Here is a first Mata based draft, maybe overkill. I will stop posting on this issue, as I do not want to spoil the thread. Thank you both for your comments. Let us focus on wishes again.

program mydoedit
    version 11.2
    mata : mydoedit()
end

version 11.2

mata :

void mydoedit()
{
    string rowvector fn
    
    fn = tokens(st_local("0"))    
    if (cols(fn)) {
        if (cols(fn) > 1) {
            errprintf("invalid '%s'\n", fn[2])
            exit(198)
        }
        if (pathsuffix(fn) == "") {
            fn = fn + ".do"
        }
        if (!fileexists(fn)) {
            errprintf("file %s not found\n", fn)
            exit(601)
        }
        stata("doedit " + fn)
    }
    else {
        stata("doedit")
    }
}

end

Best
Daniel

[Alfonso Sánchez-Peñalver]

As mentioned in another post, I also wish:

That summarize would store all the statistics for the different variables passed in varlist in a matrix so we could store them and access them throughout the code. This could become tricky with the by option.

[Nick Cox]

Daniel: Feel free to start a new thread, quoting as much of earlier posts as seems needed.

[Nils Enevoldsen]

I have a meta-request. I'd like a feature request forum. Or, better yet, for StataCorp to create a UserVoice account or a public issue tracker to solicit feedback like this.

The current method is haphazard. We have multiple threads (e.g. here's another Stata 15 thread), and new threads are started with new versions, despite 90% of the ideas in older threads still being applicable. It's difficult to search these locations, causing duplication of requests. Discussion about individual features can derail the thread. Visibility is poor, causing good ideas to receive relatively few eyeballs. There's no good way to do +1s, since people seem to mostly ignore the "like" feature on these forums. If such a forum or tracker were created, it would not be too difficult to populate with ideas collected over the years in these threads.

[Guest]

+1 to Nils Enevoldsen
Very good point!