[wbuchanan]

I would like all of the extended macro functions and expansion operators to work with macros returned in e()/r() without having to evaluate them first.

[Pete RT]

As stated here [1], it would be great if Stata 19 would be ported to GTK4 on Linux/Ubuntu and thus would natively work on Wayland. Currently Stata on Linux/Unix still uses the outdated GTK2 tool kit and, thus, still has to rely on XWayland, which already causes some problems.

[1] https://www.statalist.org/forums/for...-on-linux-unix

[Asjad Naqvi]

A request for quality-of-life improvement in dofiles:
As scripts get larger and larger, and loops, nested loops, and other blocks are added, it becomes harder to track ending brackets. One can of course collapse blocks and check, BUT it will be AMAZING if, brackets could be color coded by level. Nested loops could be darker or lighter shades of the root brackets.


It is already possible with some IDE. Would be great to have it builtin

[Bijesh Mishra]

Stata graph should not pop up like new window tabs. Can it be imbedded in the output window? that would be such a great addition to Stata. Print the code and the result underneath the window.

[Clyde Schechter]

Re #229: Different strokes for different folks. If StataCorp chooses to implement what is asked there, I hope that it would be optional and that the current behavior remains available. In my work flow, it is very convenient to have the graphs in a separate window. That way they don't scroll off the top of the screen and disappear with later commands. Also, the ability to have multiple graphs open in separate tabs within the graph window would be lost. For me, it would be very inconvenient to have the graphs show up in the Results window.

[Fahad Mirza]

Request to make egenmore and ereplace on SSC as a part of egen. Both authored by Nicholas J. Cox

[Alex Weckenman]

I would find it helpful if Stata does not hide the "time" element of a datetime when importing a datetime column where the time element is sometimes missing.

When importing a dataset from, e.g., a spreadsheet, with a variable that is sometimes a datetime and at other times just a date without a time element, the variable will be imported to look like just a date, but the underlying data will retain the time element. This leads to confusing results, for example when you compare inequalities related to multiple date variables. This can be fixed using the "cofd" command, which will assign a time of midnight for all of the dates missing the time element, and then allows sensible comparisons between datetime variables. However, it is quite difficult to identify this issue because the time element is completely hidden in Stata on import and you might not think to second guess whether the original data sometimes had a time element.

Here is a related forum thread, which I found helpful in my work when I came across this issue: https://www.statalist.org/forums/for...importing-time

[Rich Goldstein]

when I use -sem- with more than one mediator, -estat teffects- produces only the total indirect effect; please add, as an option, the ability to get each separate indirect effect

[wbuchanan]

Allow the version statement to affect the behavior of Mata as well. For example:

version 15: run someMataCode.mata
version 15: lmbuild libmymata

Should compile the Mata library under version 15 of Stata. Currently it does not seem to work this way and can make it a bit challenging to maintain code designed to work with older versions of Stata.

[Craig Kerr]

It would be great to have the functionality of missings dropobs in import excel. I often get excel sheets of data from others that import all empty rows as data and it takes a long time to load. I wonder if it's possible for import to skip over empty rows in a way that would be faster than loading and subsequently dropping empty rows using missing dropobs.

[wbuchanan]

I would like for mata describe, mata which, and other similar mata commands to return values in some way. For example, if I want to test whether a specific mata library is included in the current mata library index, it would be great to be able to do:

mata:
x = mata mlib index
end

And have the list of Mata libraries that are searched stored in x. Similarly, if mata which could return the library or source code name where the function/object exists that would also be really useful.

[Chul-Kyoo Jung]

The Do-file Editor should incorporate more user-friendly editing features, such as 'find and replace', similar to those found in other text editors such as R studio.

[Leonardo Guizzetti]

The Do-file Editor should incorporate more user-friendly editing features, such as 'find and replace', similar to those found in other text editors such as R studio.

This exists already.

[wbuchanan]

It would be great to have a try/catch and/or capture function in Mata as well.

[Felix Kaysers]

Easier reporting of regression estimates in putdocx to make reproducible research. For example, if I run three regressions, and I want to report the estimates of a variable for all three regressions, I need to store the results for all three regressions, then I need to restore the estimates for the second regressions, write my putdocx text and then restore the results for my first regression and then use putdocx text again. This becomes tedious and overly complex, if I want to report the results over different models. It would be nice to have a functionality similar to R, where I can run a regression model, store it and then refer to different model when writing up results.