[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: Date/time and "log close" after error message
That's a question for StataCorp.
There is an easy way to find out about undocumented
. whelp undocumented
The problem is the _non_documented commands.
> Sent: 02 December 2003 20:20
> To: email@example.com
> Subject: RE: st: Date/time and "log close" after error message
> At 08:07 PM 12/2/2003 +0000, you wrote:
> >Turn on the trace at the previous line!
> The problem arose for me in a situation where there were a
> series of
> -foreach- and/or forvalues loops, some nested. In order to
> turn on trace at
> the "previous line" I had to find where the crash occurred.
> So I had many
> (10?) minutes of watching traces display on the screen
> before the error was
> shown. If there were an indication of the failure line, I
> could then go to
> the "previous line" and save much time.
> >As the last trace could presumably be kept in memory,
> >this does sound possible, except that
> >I guess that the trace would have to
> >be displayed after a crash (not a problem,
> >in my view).
> >So you could have
> >. set lasttrace on
> >meaning that you get a more detailed
> >report on a crash.
> >Is this possible?
> I don't know, but perhaps Stata has an undocumented command
> to allow one to
> come up with this information?
* For searches and help try: