# st: Re: symbol for ">=" in graph text box

 From Kit Baum To statalist@hsphsun2.harvard.edu Subject st: Re: symbol for ">=" in graph text box Date Wed, 15 Oct 2008 07:50:09 -0400

< >

Quite so. "Upper ASCII" or "eight-bit ASCII" characters beyond decimal 127 differ depending on operating system and 'code page'. I found yesterday that comparison of the ISO-8859-1 ("Latin-1") and MacRoman character sets shows that some characters are in the former and not the latter, and v.v. The \leq, \geq symbols are present in MacRoman and missing from ISO-8859-1 (which is a subset of "Windows-1252", proprietary to M\$ as MacRoman is proprietary to Apple). Likewise the superscript 2 and 3 symbols are in ISO-8859-1 but missing from MacRoman.


IMHO publication-quality graphics will only be available in Stata when Stata allows for Unicode specifications, that is, mixing character sets in a graph (as one can in a LaTeX text document) to include such things as Greek characters, math symbols, accented characters of all sorts in graphs, without resort to hacks such as tweaking the PostScript files. This is also important for output in the Results or SMCL log: if you're labelling variables or outcomes using accents, you want to be able to see them that way (and not just the common French/ Spanish/German accents). Presently the situation w.r.t. output tables is more readily handled, as programs such as -estout- allow for the inclusion of any LaTeX directive when producing a TeX table.

Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html

On Oct 15, 2008, at 02:33 , Nick wrote:




As I understand it, ASCII characters up to 128 really are standard. From 129 up, it depends on your set-up. What is visible and easy to one person will be invisible or impossible to another on a different system. Stata cannot standardise what is beyond Stata.


I am among those who directly or indirectly have not spelled out the second statement sufficiently in various writings. At some point we should hit the -asciiplot- help to make this clearer. However, I am also among those who do not wish to make categorical statements that may be quite wrong for set-ups I can't access (meaning, most).

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/