Any character? I doubt it. But you can I think get away with anything
that -char()- recognises. Here's a dopey example.
. gen mla2 = cond(foreign == 0, "`=char(216)'", "`=char(247)'")
. graph matrix headroom trunk weight length, ms(none) mla(mla2)
mlabpos(0)
Mind you, a quick look at the results of -asciiplot- from SSC doesn't
suggest many symbols good for graphs that aren't accessible otherwise.
See also
SJ-5-4 gr0023 . . . . Stata tip 27: Classifying data points on scatter
plots
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . N.
J. Cox
Q4/05 SJ 5(4):604--606 (no
commands)
tips for using separate, gray-scale gradation, and text
characters as class symbols to classify data points on
a scatter plot
which is freely visible at
http://www.stata-journal.com/sjpdf.html?articlenum=gr0023
Nick
[email protected]
David Airey
I always seem to forget that trick. I suppose one could use any
character from a symbol font too.
On Mar 31, 2009, at 7:59 AM, Nick Cox wrote:
> As Martin says, -mcolor()- never takes anything other than something
> indicating a single colour.
>
> So, either you generate the individual graphs using -scatter- and then
> -graph combine-, or you distinguish genders in another way.
>
> Here's another easy solution, if not to your problem:
>
> . sysuse auto, clear
> (1978 Automobile Data)
>
> . gen mlbl = cond(foreign == 0, "o", "+")
>
> . graph matrix headroom trunk weight length, ms(none) mla(mlbl)
> mlabpos(0)
>
*
* 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/
