st: reading value labels -- second and third largest values

[Jakob Pohlisch <mailjp@gmx.net>]

Hello List,

i have some short questions concerning user written ado-Files. Iam 
writing my own little program here and this list has been an awesome 
help for me so far. Thanks to everyone involved!
But now i finally have some problems which a search in the list can't 
solve...or i just cannot find the answer.

1. The command summarize is computing the 4 largest and smallest values. 
Unfortunately the command is not storing them for the reason that it is 
a build in command i cannot manipulate it to do so. To get this values 
from a dataset is not that hard. i would just sort the values and take 
the values of the observations _N, _N-1 and so on. But if i would like 
to use the bysort "command" i would have to get the values of 
"subsamples" of my datatset. This is not to hard if i use the preserve 
and restore commands. But these commands unfortunately are VERY 
time-consuming. I have really big datasets on hand and would like to 
obtain those values without the preserve and restore commands. Any idea?

2. I managed to get the levels of a variable with "levelsof `var', 
local(levels)". Now i would like to obtain the value labels of those 
values if they exist. This will help me displaying my results.

Thank you so much in advance and sorry for all the text and my bad 
writing ;-)

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