Hi Gró,
I worry that there is something wrong with the data itself. Are there odd
text strings in the data of some the variables? If you could run the
imputation in R instead of AmeliaView, we could get a more informative
error message and maybe have a better guess as to what is going on.
Another guess is that your data has some variables that are nearly linearly
dependent. That is, you might have a bunch of scores and the sum or mean of
those scores all included in the data. Maybe you could try a different
strategy---put almost all of the variables into the ID vars and then see if
it runs and then add variables one at a time to see the culprit.
Cheers,
Matt
On Wed, Oct 8, 2014 at 8:35 AM, Gró Einarsdóttir <gro.einarsdottir(a)psy.gu.se
Hi Renzo!
Thank you so much for your quick reply and helpful comments! I really
appreciate it!
I examined the data using bivariate correlations for all variables and
found that several variables had correlations above .80. So I excluded
these variables from the analysis by changing them to ID variables and ran
the imputations. This seemed to help because I got the information that the
imputation was successful. I got five excel files as output.
When I wanted to do the same thing again with my other datasample (I have
3 datasamples with similar variables) the program started crashing again. I
tried a couple of times by excluding more variables but the program kept on
crashing.
In order to understand more about the problem I went back to the start,
and tried conducting the same analysis again with the first data sample
(and excluded the same variables again). When I did this the program
continued on crashing.
I really do not know how to move on from here. Any suggestions?
All the best
Gró Einarsdóttir
PhD student
Gothenburgs University
------------------------------
*Från:* amelia-bounces(a)lists.gking.harvard.edu [
amelia-bounces(a)lists.gking.harvard.edu] för Renzo Carriero [
renzo.carriero(a)unito.it]
*Skickat:* den 7 oktober 2014 14:29
*Till:* amelia(a)lists.gking.harvard.edu
*Ämne:* Re: [amelia] Crashing when pressing impute
Hi Gro,
I had a similar problem, but then I realized I made a mistake with the
imputation model. There were some near collinear variables, if I remember
well. As soon as I discarded them, the imputation ran well. My advice,
however, is to use Amelia in R, it is not that difficult, it is sufficient
to copy and edit some lines from the documentation.
Best
Renzo Carriero
University of Turin, Italy
Il 07/10/2014 13.12, Gró Einarsdóttir ha scritto:
Dear Amelia mailing list,
First, let me say that I really appreciate all the help I can get with
this problem. Thank you for providing this great service for free and for
all your hard work. Now to the issue.
Being disappointed with the Multiple imputation function in SPSS I was
very excited when I came across Amelia II. Because I am not familiar with
working with R I was even more excited learning about the Amelia view
function.
So I installed the latest version of R and Amelia II, view version and
uploaded my survey data.
I have N=611 and 87 variables. There is about 25% missing values. I am
able to open both R and Amelia view without problems and upload my data.
However, when I press impute this error message below appears and the
program crashes
*R for Windows terminal front end has stopped working. A problem caused
the program to stop working correctly. Windows will close the program and
notify you if a solution is available.*
This error message is very vague and I do not know what the problem is.
I have tried the program on a different PC, I have installed the amelia
package again and tried using different data samples.
I also got some help from the open science frameworks statistics help. The
statistics consultant there was able to run some multiple imputations using
my data so the problem does not seem to be with the data sample.
Any ideas about what the problem might be and how I can move forward?
All the best
Gró Einarsdóttir
PhD student
Gothenburgs University
--
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 listAmelia(a)lists.gking.harvard.edu
To unsubscribe from this list or get other information:
https://lists.gking.harvard.edu/mailman/listinfo/amelia
--
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(a)lists.gking.harvard.edu
To unsubscribe from this list or get other information:
https://lists.gking.harvard.edu/mailman/listinfo/amelia