Dear Christopher,
I have multiply imputed two data-set (let's say Africa1 and Africa2).
Now I run 1 regression (let's call it: reg1) using the imputed data from
Africa1 and 1 regression (let's call it: reg2) using the imputed data from
Africa2. For these 2 regressions I use Zelig that
Seems you're after the pooled results. Would the following work?
library(Amelia)
library(Zelig)
library(xtable)
data(africa)
m = 10
imp1 - amelia(x = africa,cs=country,m=m)
imp2 - amelia(x = africa,cs=country,m=m)
lm.imputed1 - zelig(gdp_pc ~ trade + civlib, model=ls,data = imp1)
lm.imputed2
That's right!
Your advice is in the right direction and with little adjustments it did
the job. However, I admit it was tricky and the result looks a bit
artisanal and needs some polishing that I will do by hand in the tex code.
Is it possible that there is no way to get nicely latex formatted
Hi,
I am glad you could get it to work. I don't really know I usually just
use xtable and any additional formatting I need done I do in my LaTeX
editor. Perhaps there isn't a nice tex format out of the box for MI data.
Once you write some nice code, you could keep reusing it or better yet
package
Dear listers,
I am running some OLS on multiply imputed data using Amelia.
I first imputed the data with Amelia.
than I run a OLS using Zelig to obtain a table of results accounting for
the multiply imputed data-sets. And I'd like to do this for various models.
Finally, I want to output all the
Hi Christopher,
thanks for your reply. Unfortunately, that's not what I am looking for. I
would like to have a table with the results of the two models (lm.imputed1
and lm.imputed2) in two separate columns.
According to stargazer syntax I should type something like:
stargazer(lm.imputed1,
Does this do what you want?
library(Amelia)
library(Zelig)
library(stargazer)
library(xtable)
data(africa)
m = 10
imp1 - amelia(x = africa,cs=country,m=m)
imp2 - amelia(x = africa,cs=country,m=m)
lm.imputed1 - zelig(infl ~ trade + civlib, model=ls,data = imp1)
lm.imputed2 - zelig(infl ~ trade +
What do you mean by results? Do you want just the estimated parameters? And
are you looking for one big table with all the estimated parameters from
all imputation runs?
Chris
On Sat, Aug 17, 2013 at 11:18 AM, Francesco Sarracino f.sarrac...@gmail.com
wrote:
Hi Christopher,
thanks for your
Oh and are you looking for just the summarized results over all the imputed
runs? i thought you wanted them from each iteration.
On Sat, Aug 17, 2013 at 11:38 AM, Christopher Desjardins
cddesjard...@gmail.com wrote:
What do you mean by results? Do you want just the estimated parameters?
And
9 matches
Mail list logo