Removing leading 0's in variable names to convert wide form to long form panel data

Hello,

I'm quite frustrated with a few different efforts I have made to convert my variables from wide-form to long-form.

I have a labour panel database that has over 800 variables for 3 waves. Variables are noted var_98 var_06 var_14. When converting to long form I get a warning that stata cannot find var_6.

From reviewing other's requests, I see that the leading 0 is an issue for stata and am trying to find ways to remove it without manually typing all the variables in my do file.

So essentially this would mean converting all var_06 = var_6. I see that some have suggested using substring and subinstr but I get mismatch errors, and am not sure if this only does apply to the variable name rather than the actual name of the variable.


foreach stub of varlist gov_06-RmnthwgAllJobCPI12_06 {
2. local newname = substr(`stub', 1, length(`stub')-1) if substr(`stub',-1,1)
> =="0"
3. rename `stub' `newname'
4.
. }

type mismatch
r(109);

Secondly, if possible I would prefer that it just applies to the entire list whenever the final two numbers are 06. Could someone let me know if they have tips on solving this?

Michelle