Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE: Will Stata/MP speed up running multiple dofiles inbatch mode?

From   Ada Ma <[email protected]>
To   [email protected]
Subject   Re: st: RE: Will Stata/MP speed up running multiple dofiles inbatch mode?
Date   Thu, 20 Jan 2011 11:57:00 +0000

Hi Alex and Neil,

I don't know if my experience is relevant as I was using Windows
version of Stata.  I am also not entirely sure if moving to MP will
make it any faster for you but if you are working in an university
environment, you might be able to borrow a copy of MP and test it out
before you buy a copy.

I have ran multiple do files simultaneously on PCs - both on SE and on
MP.  The processing time may or may not get longer for all processes -
it all depends on how much resources is still left on the PC - on a
Windows PC you can check the Task Manager and look at the Performance
tab to see if you are already using 100% of the computer's processing
power - there must be equivalent on a Mac / Unix so checking that out
should help.  If the time used by each of your batch run is already
getting longer, it seems to me that your PC doesn't have much
resources left and I am guessing that moving to better software isn't
going to do much help becoz the hardware is already stretched to

I was doing bootstrapping - I didn't use the bootstrap command but I
wrote my own loop.  On a very fast computer and if I let it loop
through it all 1000 times it would have taken 7 days.  By starting
several processes at the same time - and seeing on the Performance tab
that the CPUs are already running at 100% at most times - I cut the
process down to 2.5 days without seeing each of the loop having been
slowed down.   So eventually I borrowed others computers and split the
loops across four computers (2 fast 2 slow) and cut the total
processing time down to 1 day.


Ada Ma
Research Fellow
Health Economics Research Unit
University of Aberdeen, UK.
Tel: +44 (0) 1224 555189
Fax: +44 (0) 1224 550926
*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index