[Chen Samulsion]

Add "replace" option to "encode" command. Several user-written commands such as rencode and sencode had accomplished that.
See http://homepages.rpi.edu/~simonk/stata/rencode.ado
http://www.rogernewsonresources.org....10/sencode.zip

[Ruzica Savcic]

It would be great to return the option of copying a Stata .gph graph to a Word file without having to save it in a different format.

[gao weixuan]

I want a command to calculate AUC after Cox regression

[Mario Ferri]

Panel Vars, better graphs, including 3D, Better Bayesian and DSGE Models.
They should simply try to copy and adapt to Stata the BEAR toolbox from ECB and Dynare toolbox for DSGE .
Current state of the art research is based on coding. Stata should adapt to the standards otherwise risks to remain to the oldstone era and out of market

[JanDitzen]

Current state of the art research is based on coding. Stata should adapt to the standards otherwise risks to remain to the oldstone era and out of market

What do you mean with the "Current state of the art research"? When developing methods, researchers always need to do the coding. Stata will (or is likely) to lack behind the frontier.

[George Ford]

Use egen to get the number of unique values of a variable. by friendly.

bys X: egen X2unique = unique(X2)

[William Lisowski]

Re #171:

ssc install egenmore
help egenmore
egen X2distinct = nvals(X2), by(X)

[Leonardo Guizzetti]

I would like a way to specify table column widths when using -putdocx-, even if it's in a way that mirrors -putpdf- by specifying relative widths.

[Justin Niakamal]

This is an extremely petty request and I know it’s just a single line of code (drop _freq), but I’ve always wanted a nogen option added to contract, the same way that you can forgo the creation of a _merge variable using the nogen option with merge. I’d like to have the option to not have a _freq variable.

[Chen Samulsion]

I think it will be very fun to devote some time to review historical wishlists:
https://www.statalist.org/forums/for...t-for-stata-14
https://www.statalist.org/forums/for...t-for-stata-15
https://www.statalist.org/forums/for...t-for-stata-16
https://www.statalist.org/forums/for...t-for-stata-17

[John Mullahy]

Perhaps consider implementing a function (say pwr(x,p)) that returns, as appropriate, particular powers of possibly non-positive arguments.

For instance,

pwr(-8,1/3)

returns -2

The function behaves just like x^p for x>0 but behaves like

sign(x)*abs(x)^p

for all x.

Such a function would allow one to circumvent this kind of result,

. gen z=-8

. gen pz=z^(1/3)
(100 missing values generated)

where what we would like is for pz to equal -2.

[Clyde Schechter]

A frequent problem seen on this Forum is the omission of variables from regression models due to a colinearity that the user does not expect/understand. Sometimes these are more or less obvious to an experienced user. Occasionally they are truly difficult to ferret out. It would be nice if estimation commands came with an option to amplify the "...omitted due to colinearity" messages so they explain which other variables participate in the offending colinearity.

[Daniel Jensen]

This might seem a bit petty, I think there there are quite a few user-written commands for data management that should become part of official Stata. For example:

• sreshape (or at least sreshape's added efficiency and capability added to the official reshape command)
• readreplace (with the bug introduced in Stata 17 fixed).
• missings
• reclink

Since new analysts usually get stuck with the data management/cleaning tasks, it would be nice to be able to hand them a do file that runs without having to install several user-written commands first.

Also, I think it's about time that some version of psmatch2 became part of official Stata (if only so someone can clean up the help file).

[Fei Wang]

I wonder if it's possible to allow -bysort- to sort in both ascending and descending orders. Just like a combination of -by- and -gsort-.

[alejoforero]

Macro list sorting in reverse order, like gsort for macros. There is a function for normal sorting only.

global unsorted= "MEX ALB CHN"
global sorted: list sort global(unsorted)
di "$sorted"
ALB CHN MEX

Would be something like:

global gsorted: list sort global(unsorted), reverse
di "$gsorted"
MEX CHN ALB

There is a workaround using MATA, but its impossible to remember:

mata : st_global("gsorted", invtokens(sort(tokens(st_global("unsorted"))',-1)'))

This would be extremely useful for loops.