There is also a 'pair' function in the cem package, which I think produces a
paired sample out of a CEM match solution.
However, I realized that pair function pairs up instances that are not necessarily
matched.
So I modified the pair.r file in the cem package, and force the pair function to pair up
instances that are matched already.
Below is what I diid and hope it helps:
pair <- function(obj, data, method=NULL, mpower=2, verbose=0){
nm <- NULL
strataID <- unique(obj$strata)
reservoir <- NULL
for(i in strataID){
####################################################
###########added by hoktay
########use only the instances that are matched
#### used to be
######### idx <- which(obj$strata==i )
####################################################
idx <- which(obj$strata==i & obj$matched==TRUE)
.....
----------------------------------------------------------------------
Message: 1
Date: Sun, 12 Feb 2012 23:44:05 -0800
From: "Ashkan Moazzez, MD" <ashkanmoazzez(a)ucla.edu>
To: cem(a)lists.gking.harvard.edu
Subject: [cem] CEM
Message-ID: <20120212234405.79195ckdlsb7deth(a)mail.ucla.edu>
Content-Type: text/plain; charset=ISO-8859-1; DelSp="Yes";
format="flowed"
Dear members of the CEM list,
I was wondering if there is a way to extract the data after you match
them with CEM? Is there an equivalent command like match.data in
matchit() for CEM?
Best regards,
Ashkan
Ashkan Moazzez, MD
------------------------------