create and call subroutine from the same .do file

[Francois Durant]

Hi,

I have a large piece of code that I would like to break down in several smaller pieces of code that I could run (or not) from the top of the .do file. I do not want to break down the main do file in several do files but rather have more control over the part of the code I am running in 1 single do file.

* Main file:
** Call 1st piece of the code: 1) Merging data
***** Some code that calls the 1st part "merging data"

** Call 2nd piece of the code: 2) regression analysis
***** Some code that calls "regression analysis
etc.


I have tried the following but I do not get any display. Anyone can point to the right direction/format?Thanks!

clear
capture program drop main
program main
    test
end

capture program drop test
program test
    di "Hello World"
end

[William Lisowski]

In the code you show, you have defined two programs but you haven't run either of them.

. main
Hello World

.

[Francois Durant]

Thanks William.When I call run "test" for instance, do I always need to run first the part:

 
 capture program drop test program test     di "Hello World" end

Otherwise, I am getting the following error:
. test
last estimates not found

Is there any way to go around the program? I do not necessarily want to run the "program ... end" first. Is there a way around it? Thanks.

[Francois Durant]

Also, I have tried to copy/paste pieces of codes (that actually work without the subroutine structure) inside "test" and I am getting the following error:

main
command main is unrecognized

Would you be able to help me understand why? Thanks.

[Francois Durant]

Any thought William Lisowski ? Thanks

[JanDitzen]

The code example works for me.

However a problem might be that test is a build in Stata program and there might be clashes in terms of priority.

[William Lisowski]

The follow-on problem in posts #3 and #4 were subsequently reposted with the addition of a reproducible example at

https://www.statalist.org/forums/for...e-in-a-do-file

which led quickly to a response, as the reproducible example in post #1 led to a quick answer in post #2, and the lack of an example in posts #3 and #4 led to a very slow response in this topic.

When this problem was reposted, along with the link in the new topic back to this topic, a link forward to the new topic should have been added to this topic.