How to fix subscripts in Stata graphs

[Tiago Pereira]

Dear Statalisters,

A previous post addressed a similar problem, but had no success in getting a solution:
https://www.statalist.org/forums/for...some-instances

In a nut shell: when we have multiple subscripts in a graph, only the first is OK. The remaining subscripts fail.

Here is an example:

clear
set seed 1234556
set obs 200
gene time = round(runiform()*7)
gene group = runiform()>0.5
gene y = cond(group==1,rnormal(100,5), rnormal(90,5))
twoway qfitci y time if group==1 || qfitci y time if group==0, xlabel(0 "Baseline" 1 "T{subscript:1}" 2 "T{subscript:2}" 3 "T{subscript:3}" 4 "T{subscript:4}" 5 "T{subscript:5}" 6 "T{subscript:6}" 7 "T{subscript:7}")

Has anyone found a solution for that? Is it a bug in Stata? I am under Stata SE 14.2 for Linux (Revision 16 Nov 2016).

All the best,

Tiago

[Sergiy Radyakin]

Dear Tiago,

this seems to execute correctly in Stata 16.0 on Windows. So it is either

• already fixed, or
• is platform-specific, or, perhaps,
• is caused by some other factor that is not captured by your posted example.

Best, Sergiy Radyakin

clear all
set seed 1234556
set obs 200
gene time = round(runiform()*7)
gene group = runiform()>0.5
gene y = cond(group==1,rnormal(100,5), rnormal(90,5))
twoway qfitci y time if group==1 || qfitci y time if group==0, xlabel(0 "Baseline" 1 "T{subscript:1}" 2 "T{subscript:2}" 3 "T{subscript:3}" 4 "T{subscript:4}" 5 "T{subscript:5}" 6 "T{subscript:6}" 7 "T{subscript:7}") name(g1) legend(rows(1))
twoway qfitci y time if group==1 || qfitci y time if group==0, xlabel(0 "Baseline" 1 "T{superscript:1}" 2 "T{superscript:2}" 3 "T{superscript:3}" 4 "T{superscript:4}" 5 "T{superscript:5}" 6 "T{superscript:6}" 7 "T{superscript:7}") name(g2) legend(rows(1))
graph combine g1 g2 , rows(2)

[Tiago Pereira]

Thank you so much for your prompt reply, Sergiy.

It seems it has to do with either Stata 14 or linux. I will be checking the same code in a different platform.

Thank you again.

All the best,
Tiago

[Michael Wiebe]

I get the same problem on Stata16 on linux, but only for subscripts.