Roman numerals inside a text string

[Esteban Jara]

Hello. Could someone give me some trick to detect roman numerals inside a text string? I have looked at the options with the regular expressions, but I can't think of any... thanks

[Nick Cox]

See https://www.stata-journal.com/articl...article=dm0054 which may help.

[Esteban Jara]

I have this data:

* Example generated by -dataex-. To install: ssc install dataex
clear
input str13 var
"VICTORIA)"    
"S"            
"UNIDAS)"      
"DE"           
"UNION"        
"JARA"         
"PARRA"        
"PUCARA"       
"CORDILLERILLA"
"LABRANZA)"    
"AROMOS)"      
"LAS"          
"IMPERIAL)"    
"ROZAS)"       
"LIDER)"       
"NUEVO)"       
"ORIENTE"      
"1)"           
"4)"           
"BARRERA"      
end

and I have this error message:

. fromroman var, gen(x)
subinstr(): 3200 conformability error
from_roman(): - function returned error
<istmt>: - function returned error
r(3200);


Perhaps the variable has some data that makes it difficult to use the command? Because I used the command on other text variables and it worked perfect.

Thank you so much

[Nick Cox]

I can’t see any Roman numerals in your example data and in any case the command does not deal with text that includes anything else.

[Esteban Jara]

Does it mean that the command does not work well if the variable does not contain any roman numerals? Ok, I can fix it by doing some sort of rustic fix to my do-file because I need that possibility to be actually considered, but the command will be extremely useful anyway. Thank you very much again!

[Nick Cox]

I can't follow #5. If a variable does not contain any roman numerals, there is nothing to do so far as fromroman is concerned, but it will throw an error.