Statalist


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

st: Re: Faster fracplot?


From   "Martin Weiss" <[email protected]>
To   <[email protected]>
Subject   st: Re: Faster fracplot?
Date   Mon, 10 Nov 2008 23:05:06 +0100

Might want to take a look at:

http://www.stata-journal.com/sjpdf.html?articlenum=gr0013


HTH
Martin
_______________________
----- Original Message ----- From: "Johan Sundström" <[email protected]>
To: <[email protected]>
Sent: Monday, November 10, 2008 11:01 PM
Subject: st: Faster fracplot?


Hi Statalisters!

I have the following problem: In a dataset with 1,2 million observations,
using a reasonably fast computer, Stata MP (two processors) 10.1, and no
other programs running (I've even kicked up a slim msconfig for this
purpose, so I can maximize memory), I want to use uvrs or mvrs in order to
examine the shape of a relation. With up to five covariates mvrs takes ca
15-20 minutes to calculate. So far so good. I then want to fracplot. This is
where my workday goes into slow-mo. Every graph takes up to a couple of
hours to complete, and is nearly impossible to tweak even the slightest
because of the time everything takes. I've tried mstyle(none), which doesn't
help. My questions is if it's possible to ask fracplot to not draw the
markers at all (I'm only interested in the regression line and 95% CI)? Or
does anyone have another alternative for me? I've tried mfp as an
alternative to splines, which is equally slow. Help, please!

Best wishes,

Johan Sundstrom
Uppsala, Sweden

*
*   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/



*
*   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index