Error Message: command makecns not defined by makecns.ado

[Rosanna Hildersley]

Hi,

I've been trying to run a bivariate logistic regression using a binary dependant and categorical independent variable. I'm receiving an error message reading "unexpected end of do-file command makecns not defined by makecns.ado." Neither the "logit" or "logistic" commands work, although linear regression "regress" does work.


I’ve tried searching for this error message but I can’t find any explanation. I understand that makecns.ado is one of the STATA programme files, and I haven’t touched this. Also, I haven’t added any personal .ado files to my copy of STATA, and it was only installed on this computer a few weeks ago.

Can anyone help with this, or explain what the error message means?

Thanks!

[Mike Lacy]

Something strange is going on here. I can't see how the -logistic- or -logit- commad would call -makecns-. Please show exactly what you typed and exactly what Stata responded.

[Rosanna Hildersley]

I've attached a screenshot with my .do file and stata code, I've never had to post here so I hope that works!

My colleague says that the code is running fine on her PC.

Thanks for your reply!

[William Lisowski]

I am going to hazard a guess that there was or is a problem with your Stata installation. I suggest you download and run the Stata Installation Qualification Tool from the Stata Support pages at

https://www.stata.com/support/instal...qualification/

Also, this is the sort of thing that Stata Technical Services absolutely can help you with, and you should not hesitate to contact them following the guidance at

https://www.stata.com/support/tech-support/

perhaps even before running the Installation Qualification Tool, if you are concerned about running it.

Added in edit: You seem to be running in a complicated computing environment, based on the location of the profile.do in the screenshot on the right. If I'm correct, and it was not you who installed Stata, you should probably start with your IT support staff. What I wrote above was predicated on the idea that you were running Stata in a relatively simple setup.