16 Apr
2021
16 Apr
'21
2:54 p.m.
Hi Antonio, for Amelia to work, there has to be enough info in the
dataset. E.g., suppose you have data in this format:
1 ?
? 1
where ? is missing and 1 is observed and there are lots of rows for each.
then you can't use the first set of rows to impute the ?. and you can't
use the 2nd set of rows to build a model to infer the first set of rows
because there is no overlap at all in what is observed. You have a more
complicated setup, but the issue is probably an issue like this.
I'd start with a much simpler case than you desire to work with and then
gradually make it more complicated/sophisticated, and you'll get a feel fro
what's going on.
Best of luck with your research
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS <http://iq.harvard.edu/> - Harvard University
GaryKing.org <http://garyking.org/> - King(a)Harvard.edu - @KingGary
<https://twitter.com/kinggary> - 617-500-7570 - Assistant
<king-assist(a)iq.harvard.edu>du>: 617-495-9271
On Fri, Apr 16, 2021 at 10:25 AM Antonio Rodriguez Andres <
antoniorodriguezandres70(a)gmail.com> wrote:
Hello Amelia users
I have tried to use the amelia package, and I got the following error. I
am dealing with a macro panel dataset for 126 countries and 16 years. I
want to impute using the entire dataset, the values of time to export, time
to import (measured in hours), and cost of exports and imports in USD.
The proportion of missingness is quite large around 45 and 65 % for the
variables. For that purpose, I use the option empri, read Amelia manual on
page 31.
am = amelia(final_datos1, m = 5, intercs = TRUE, empri =
.01*nrow(final_datos1), cs = "country", ts = "year")
After running this, it gives me the error
Warning message:
In amcheck(x = x, m = m, idvars = numopts$idvars, priors = priors, :
The variable year is perfectly collinear with another variable in the
data.
It seems to me that this is due to the fact that for the combination of
country and year, there are missing values for all variables
Any help is appreciated
--
Member, Editorial Committee, *The Economic and Labour Relations Review* (a
SAGE journal)
http://elr.sagepub.com/
Member, Editorial Committee, African Journal of Economic and Management
Studies
http://emeraldgrouppublishing.com/products/journals/editorial_team.htm?id=a…
https://www.researchgate.net/profile/Antonio_Andres (Research Gate
profile)
--
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 send an email to amelia-leave(a)lists.gking.harvard.edu
To unsubscribe from this list or get other information:
https://web.lists.gking.harvard.edu/mailman/listinfo/amelia