Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Re: Stata Intercooled and memory requirements with micombine command


From   "Rodrigo A. Alfaro" <ralfaro76@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st: Re: Stata Intercooled and memory requirements with micombine command
Date   Fri, 16 Jun 2006 10:14:11 -0400

I forgot to mention that the command miest can deal
with multiple datasets instead of one-dataset with
multiple imputations, as micombine does.

Then you should get the same using less memory.

Rodrigo.

----- Original Message ----- 
From: "Suzy" <scott_788@wowway.com>
To: <statalist@hsphsun2.harvard.edu>
Sent: Friday, June 16, 2006 9:52 AM
Subject: Re: st: Re: Stata Intercooled and memory requirements with 
micombine command


Thank you so much for the recommendations and tips, Rodrigo.

Rodrigo A. Alfaro wrote:

>Suzy,
>
>Hard problem if you cannot increase memory RAM.
>
>The command -ice- seems that generates the imputed dataset in
>one dataset. This generates a final dataset which is m*initial_size.
>
>You can generate just 1 imputed set in several trials:
>
>forvalues i=1/10 {
>    ice x1 x2 x3 using imp`i', m(1)
>}
>
>Then you can use miest which is available at Gary King webpage
>(http://gking.harvard.edu/amelia/amelia1/docs/mi.zip) .
>
>Moreover you can take a look of his software Amelia
>http://gking.harvard.edu/stats.shtml
>
>I hope my suggestion helps you.
>Rodrigo.
>
>
>----- Original Message ----- 
>From: "Suzy" <scott_788@wowway.com>
>To: <statalist@hsphsun2.harvard.edu>
>Sent: Wednesday, June 14, 2006 2:06 PM
>Subject: st: Stata Intercooled and memory requirements with micombine
>command
>
>
>I am attempting to conduct statistical analyses with m=10 imputations
>via the ICE and micombine commands. I've generated the imp.dta file just
>fine and can successfully upload it into Stata (slow ~ 1-2 minutes). I'm
>using the micombine command to generate a simple regression output -
>this works just fine with m=5 imputations. Stata cannot and/or will not
>run the analysis on my PC with m=10 imputations.
>
>I have Stata intercooled - version 9.2 fully updated.
>
>My PC has 512mb RAM with virtual memory option. I've tried my best to
>optimize my PC performance settings, though I'm not sure if my settings
>are truly optimizing performance.
>
>In Stata, I've adjusted the memory for 600m, matsize 800, dropped all
>unnecessary variables and compressed the rest. In other words, I "think"
>I've maximized Stata intercooled performance.
>
>I ran a describe, short on the imputation dataset:  imp.dta:
>
>observations: 231,044
># var: 533
>Size: 268,704,172 [57.3% memory free]
>
>
>I have never had any performance problems for pictures, graphics,
>videos, movies, or any analysis with Stata.
>
>I would appreciate any help on optimizing my PC and Stata intercooled
>performance for analysing m=10 or more imputations.
>
>I sincerely hope I'm not doomed to needing to increase RAM or update to
>Stata SE. Those are not viable options for me at this time!
>
>Thanks  for any help!
>
>Suzy
>*
>*   For searches and help try:
>*   http://www.stata.com/support/faqs/res/findit.html
>*   http://www.stata.com/support/statalist/faq
>*   http://www.ats.ucla.edu/stat/stata/
>
>
>
>
>


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index