[George Ford]

I have a lot of trouble with dates. I wonder if a pull-down application could be created that let you select pieces of a date and tell Stata what the pieces are, and have Stata show how it interprets it. Then provide the command in the window.

[Maarten Buis]

So you are thinking of a menu for creating a Stata date variable from a string variable with dates as strings. You would first specify the name of the new variable and the name of the string variable. Than the menu takes the first observation from the string variable, splits it up in parts, shows those parts and below each part there is a dropdown menu where you can select if that part is a second, hour, day, week, month, quarter, or year. Maybe allow the user to select another observation in case the first is a missing value or some other thing that still needs to be cleaned. To show how Stata interprets it, you would also have to specify the display format, but that seems possible as well. After you have made all those selections Stata could show the result for that observation. When pressing OK you it creates and executes the commands and you can see those in the results and history window.

That seems doable and useful.

[Giovanni Russo]

An expansion of the SEM and GSEM suite, the 3-step approach for LCA and other estimation methods that are less sensitive to deviations from normality. Mplus and LatentGold offer a wider set of options

[George Ford]

Exactly, Maarten. Could have a buttons that say "monthly, daily" etc... It could get really fancy with links to business calendars and so forth, but I'd be happy for starters with what you describe. As you say, with all the complicated things Stata can do, it does not seem the task is all that difficult and would be exceedingly helpful. I've spent hours wrestling with weird date formats.

[Aliyu Abdulrauf]

Dear Mr George Ford,

i am a new user to the stata, and after installing i discover that the review and variable fields are hidden what can i do to make it visible?

thank you sir

[Phil Schumm]

Please add the option of making interactive visuals.

You might want to take a look at Stata's new support for Jupyter notebooks. There are several excellent packages in Python for creating interactive graphics, e.g., Bokeh and Altair. With just a couple lines of code, you can create a rich, interactive plot from a Stata dataset (or results generated in Stata) that can then be shared or published on the web so others may interact with it too.

[Yevgeniy Feyman]

I'd love to see something incorporating the recent DID work on staggered treatment assignment.

[Michael Foreman]

this may be possible already, but i haven't found a way... so i wonder if they could add the option to extract tabulated or regression results into word or xlsx (not csv) files with an opportunity to customize the output in Stata (rename columns, use labels or variable names, format the numbers, change font etc.). This can be done to some extent with user-written functions, but hey all have some shortcoming's and dont exactly this.

[Maarten Buis]

That is what the rewritten table command does in 17

[Phil Schumm]

Indeed; this is now a done deal (https://www.stata.com/manuals/tables.pdf).

[Mario Ferri]

Time varying parameters!!!

[Bert Lloyd]

1. Official or semi-official plugins for popular text editors (Atom, Notepad++, Sublime, etc) -- there are some nice user-written contributions but official versions (or still user-written but with active help from StataCorp) with one-click installation would be a huge improvement. Even just providing a single place to look for authoritative information would be great. (First posted to Wishlist for Stata 14, so I suppose I should limit my expectations.)
2. Add a command similar to rsync that copies/overwrites only if newer (or as new options for copy, or just package rsync into Stata somehow)
3. Integration with scons, or somehow make Project Manager more like scons/make (in the spirit of the excellent project from Robert Picard )

[I-Chin Wu]

I hope Stata 18 could provide official support for Firth’s penalized likelihood in logistic regression and Cox regression. This method is useful for addressing problems of separation and monotone likelihood in the fitting process.

References:
Heinze G, Schemper M. A solution to the problem of separation in logistic regression. Stat Med. 2002;21(16):2409-2419.
Heinze G, Schemper M. A solution to the problem of monotone likelihood in Cox regression. Biometrics. 2001;57(1):114-119.
Firth D. Bias reduction of maximum likelihood estimates. Biometrika 1993;80(1):27-38.
https://cemsiis.meduniwien.ac.at/en/...th-correction/

[William Lisowski]

Note that the output of search firth finds the community contributed firthlogit command at SSC. Your reference to "official support" suggests that you are aware of this. but I mention this for others who become interested and may not think to check SSC.

[I-Chin Wu]

Currently, separation in logistic regression could be addressed by community-contributed firthlogit command. However, monotone likelihood in Cox regression could not be resolved by STATA.