Dear Xiaobo,
For your first question, regarding updating to the latest version of Amelia II,
it looks like you are using R 2.5 (even if you are using the AmeliaView
version of Amelia, your most recently installed version of R on your
machine was version 2.5). Therefore, R is finding the last version of
Amelia created under R 2.5.
If you update R to the current R 2.7.0, and then type in exactly the same
install.packages() command you used, you will get the most recent version of
Amelia.
For your second question, set the argument p2s=2, which will give more screen
output and help us understand what is going on in your EM chain. Repost the
problem with the new screen output, or send it to me directly, and I should
have a more accurate diagnosis.
For your third question, this depends on whether you are using the GUI
AmeliaView, or calling Amelia from R directly. Which are you doing?
regards,
James Honaker
> Hello all,
>
> I have three questions concerning Amelia II, and I would greatly appreciate
> any information you may provide.
>
> First, I have been trying to install the updated version of Amelia II
> (1.1-29) onto my computer today, but I keep getting the older version
> (1.1-26). Could someone please point me to the right direction? Here is the
> R command I used, and results shown in my R program:
>
> **********************************************************************
>> install.packages("Amelia",repos="http://gking.harvard.edu")
> trying URL
> 'http://gking.harvard.edu/bin/windows/contrib/2.5/Amelia_1.1-26.zip'
> Content type 'application/zip' length 2148370 bytes
> opened URL
> downloaded 2098Kb
> **********************************************************************
>
> Second, I am trying to do some diagnostics on the imputed datasets. However,
> the program is always idle at iteration 18 of the third imputed dataset. I
> can tell that that the CPU of my computer is not working at at full speed,
> as it drops down to 1% or 2% as opposed to 100% previously. I have tried to
> modify the imputation code several times and this problem always exists.
> Note that I don't have any problem when I imputed the same dataset with
> fewer variables. Is this a bug?
>
> Third, it seems that I have to conduct diagnostics immediately after the
> imputation. Is there a way to call the imputed dataset and do diagnostics on
> a later time without imputing the datasets all over again?
>
> Thanks a lot in advance for your help.
>
>
> Best,
> Xiaobo
> -
> Amelia mailing list served by Harvard-MIT Data Center
> [Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
>
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hello all,
I have three questions concerning Amelia II, and I would greatly
appreciate any information you may provide.
First, I have been trying to install the updated version of Amelia II
(1.1-29) onto my computer today, but I keep getting the older version
(1.1-26). Could someone please point me to the right direction? Here is
the R command I used, and results shown in my R program:
**********************************************************************
> install.packages("Amelia",repos="http://gking.harvard.edu")
trying URL
'http://gking.harvard.edu/bin/windows/contrib/2.5/Amelia_1.1-26.zip'
Content type 'application/zip' length 2148370 bytes
opened URL
downloaded 2098Kb
**********************************************************************
Second, I am trying to do some diagnostics on the imputed datasets.
However, the program is always idle at iteration 18 of the third imputed
dataset. I can tell that that the CPU of my computer is not working at
at full speed, as it drops down to 1% or 2% as opposed to 100%
previously. I have tried to modify the imputation code several times and
this problem always exists. Note that I don't have any problem when I
imputed the same dataset with fewer variables. Is this a bug?
Third, it seems that I have to conduct diagnostics immediately after the
imputation. Is there a way to call the imputed dataset and do
diagnostics on a later time without imputing the datasets all over again?
Thanks a lot in advance for your help.
Best,
Xiaobo
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Is there a cut-off for rate of missingness, past which we should
employ other methods (i.e. Not Amelia 2)? Or does it depend on the
diagnostic results?
More specifically, if my imputations:
a) don't give me error 34 (which says there is not enough data to do
imputations
properly) and;
b) my diagnostics seem kosher (distributions of imputed/actual
observations overlap nicely, there is convergence, etc.),
can I relax about the rate of missingness in the original data?
Simply: I got a time-series, cross-sectional dataset. 10 years, 50
countries. 6 independent vars. Of the 6, 3 have 65% missingness. Yet,
these 3 independent vars with 65% missingness have significant
relationships with the rest of the vars, and Amelia 2 was able to give
me a decent-looking imputation. [I can offer the misschk results from
Stata if necessary to answer this question.]
Is there a cut-off in the fraction of missingness past which I must
worry? Or Amelia would have already told me so?
King also mentions that upping the imputations (to, say, 10) can help
deal with higher rates of missingness. Something I should do just to
make sure?
You can also direct me to somewhere in the literature where you think
this is specifically addressed. Thanks much.
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hi all,
I ran into a problem when I tried adding range prior. I added the minimum and maximum values of
the prior and pressed OK, but got this error message "Please enter a mean value." It doesn't help
even if I add distribution prior beforehand. This did not happen when I ran AmeliaII about a month
ago. Can someone help? Many thanks.
All best,
Qian Guo
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hello all,
We just patched up a bug in the handling of "ordinals" in Amelia. The
"ords" argument forces imputations into the ordinal categories, but
this transformation had a bug in it, resulting in some incorrect
imputations. The problem was limited only to those variables marked as
"ords" so the other imputed variables would be fine. The version of
Amelia (1.1-29) will be up on CRAN shortly, but until then, you can
use the following command to update your version of Amelia:
install.packages("Amelia", repos="http://gking.harvard.edu")
The fix should resolve any issues with problematic ordinal imputations
such as unimputed variables and/or strange imputations. Note that this
transformation is usually unnecessary for most analyses and is useful
when the variable will be a dependent variable in the analysis stage.
See the Amelia manual for more information.
Please let us know if you have any problems.
regards,
the amelia team.
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Dear Developers
I have been a strong advocate of Amelia II. To date I have taught
four missing data workshops in both the US and Europe demoing Amelia
II. Two my workshops had a hands on component where I made the
participants do their own imputations. My experience teaching Amelia
II has pointed out one feature that is badly missing from the GUI. A
random generator seed setter. It would be really nice if we could set
a seed in the GUI if we wanted to so we could get the same imputations
on every computer in the classroom. It would also be very nice in
terms of logging replication information of which I am a very strong
proponents of (mainly due to the efforts of Gary King himself). But
right now (as far as I know) appropriate logging facilities through
the use of random number generator seeds are only available to command
line users of Amelia II. Additionally, some easy to use logging
function that logs the R syntax of every step taken in the GUI would
also be beneficial.
Please include these feature in a future release of Amelia II.
Thanks.
Levi Littvay
Central European University
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hello,
I am trying to impute missing values for a dataset with 1752
observations. I've tried several times with different combinations of
variables included in the imputation, and the problems described here
occur in exactly the same way every time.
First, there are several variables for which the imputation does not
occur. That is, for a handful of variables, there are the same number of
missing values in the output datasets as there are in the input. This
only happens for a handful of the variables, but it is the same ones
every time. I've looked at these variables both in their original format
(stata) and in the "summarize" feature in AmeliaView, and can't detect
anything different about them compared to all the other variables.
Second, there are a few other variables that I assigned to have an
ordinal output but for which the output is a near-zero number (2e-17).
Again, it is the same variables every time. This problem seems less
significant, as I can just change these to zeros.
Any suggestions on what might be causing this or what I can do about it
would be welcome.
Thank you,
Steve Shewfelt
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hi Amelia folks,
I have a question about imputing for highly skewed distributions. My data have missing test scores
for 540 out of 709 observations in a lower-achieving subgroup and no missing scores for the other
students (about 129,000 in size). The problem is that the subgroup has a highly left skewed
distribution while the overall distribution is highly right skewed.
I set priors by giving the mean (16.07 out of 40, versus 31.68 for the other students) and
standard deviation (9.30 vs. 7.20) of the subgroup. But would it still be a problem that the
subgroup distribution is highly left skewed (and so unlike the rest of the population)? Thanks a
lot.
All very best,
Qian Guo
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
Hello,
I am writing this because I had some problem to run "Amilia II" with missing
data. I cannot get the output.
Message says, "Amelia Error Code: 34 The number of observations in
too low to estimate the number of parameters. You can either remove
some variables, reduce the order of the time polynomial, or increase
the empirical prior. You have received an error. You can close this
window and reset various options to correct the error."
What do I have to do? Do you have any suggestions?
It's very hard to figure it out. Thank you for your information!
KyungSook
******************************
KyungSook Lee
3010 Staten Ave. #12
Lansing, MI 48910
leekyun3(a)msu.edu
******************************
Hello
I reported this error message a few weeks back (search for subject "Hard
to trace error message"). Now, with a different (improved, I hope) data
setup I am still getting stuck at this same point. Amelia runs through
the first MI step and then aborts with the message shown below.
What can I do about it? What could be the problem? (Whom could I provide
more information about the error/data to locate the problem?)
Thank you very much for your help.
Best regards,
Marcus
PS. Having N=2441, I tried different values for empri: NULL,50,100,200.
> am <- amelia(mdi,m=5,p2s=2,idvars=ids,noms=noms,ords=ords,collect=FALSE,
+ outname="Routput/imputed", write.out=TRUE,empri=NULL)
amelia starting
beginning prep functions
running bootstrap
-- Imputation 1 --
setting up EM chain indicies
1(1705) 2(1523) 3(924) 4(285) 5(55) 6(4) 7(1) 8(0)
Error in unsubset(x.orig = prepped$trans.x, x.imp = ximp, blanks =
prepped$blanks, :
subscript out of bounds
> traceback()
2: unsubset(x.orig = prepped$trans.x, x.imp = ximp, blanks = prepped$blanks,
idvars = prepped$idvars, ts = prepped$ts, cs = prepped$cs,
polytime = polytime, intercs = intercs, noms = prepped$noms,
index = prepped$index, ords = prepped$ords)
1: amelia(mdi, m = 5, p2s = 2, idvars = ids, noms = noms, ords = ords,
collect = FALSE, outname = "Routput/imputed", write.out = TRUE,
empri = NULL)
--
Marcus M. Dapp | PhD student | ETH Zurich | www.ib.ethz.ch/people/mdapp
Prof. Thomas Bernauer, International Relations | www.ib.ethz.ch
On the shoulders of giants? http://science.creativecommons.org
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia