Hi James and Matt,
Thanks for the helpful suggestions. I really appreciate it.
-Isaac
On Wed, Feb 20, 2013 at 10:27 PM, Honaker, James <jhonaker(a)iq.harvard.edu>wrote;wrote:
Isaac,
As Matt said, it's not really possible to know how many iterations will be
necessary in any particular (bootstrapped) dataset.
However, it can be useful to set the amelia option "p2s=2" for a rough
metric. The p2s stands for "print to screen" and changes the amount of
output given (0=minimal, 1=normal, 2=more detailed). At level 2, you will
see in parentheses after each iteration number, how many parameters "have
not converged" by which I mean, how many parameters in the imputation model
changed significantly between the last two rounds. Convergence is reached
when none of the parameters are moving beyond a small tolerance, that is,
this number reaches zero.
So it behaves a little like a "progress bar" but like most progress bars,
does not actually move at constant linear rate; a lot of your time can be
spent waiting for the hardest parameters, just like many progress bars can
often frustratingly spend most of it's time tantalising you at 90%
completed.
As a diagnostic tool, if when you start looking at this, you find that
much of your time is spent waiting for one or two last parameters to
converge, you might add (or adjust) a small empirical prior (see the option
empri).
James.
--
James Honaker, Senior Research Scientist
//// Institute for Quantitative Social Science, Harvard University
________________________________________
From: amelia-bounces(a)lists.gking.harvard.edu [
amelia-bounces(a)lists.gking.harvard.edu] On Behalf Of Matt Blackwell [
m.blackwell(a)rochester.edu]
Sent: Wednesday, February 20, 2013 5:31 PM
To: Isaac Petersen
Cc: Amelia Mailing List
Subject: Re: [amelia] Progress Bar in Amelia
Hi Isaac,
Unfortunately not. It's difficult to know before an imputation starts how
long it will take because this depends on the data itself. That is, there
isn't a fixed number of iterations per imputation.
Hope that helps!
Cheers,
matt.
~~~~~~~~~~~
Matthew Blackwell
Assistant Professor of Political Science
University of Rochester
url:
http://www.mattblackwell.org
On Wed, Feb 20, 2013 at 5:24 PM, Isaac Petersen <dadrivr(a)gmail.com<mailtomailto:
dadrivr(a)gmail.com>> wrote:
Is there any way to get a progress bar for the imputation progress in the
command line version of Amelia? I'd like to be able to estimate how long
the whole imputation process will take. For loops, one can use the
following for a progress bar:
pb <- txtProgressBar(min = 0, max = 999999, style = 3)
for (i in 1:999999){
setTxtProgressBar(pb, i)
}
Is there an option to get a progress bar for a single or multiple
imputations in the command line version of Amelia?
Thanks!
--
Amelia mailing list served by HUIT
[Un]Subscribe/View Archive:
http://lists.gking.harvard.edu/?info=amelia
More info about Amelia:
http://gking.harvard.edu/amelia
Amelia mailing list
Amelia@lists.gking.harvard.edu<mailto:Amelia@lists.gking.harvard.edu>
To unsubscribe from this list or get other information:
https://lists.gking.harvard.edu/mailman/listinfo/amelia