Matchit September 2016

matchit@lists.gking.harvard.edu

2 participants
2 discussions

Re: [matchit] Matchit Digest, Vol 133, Issue 1

by Andreu Ferrero

Hey, I found this commads on the net to do what I need... but I am not good enough at programing. There are any easier option? *** Hi Sharon, I was just working yesterday to reconfigure my data so that it includes matched pair identifiers. I did not take my data to the matched pair level of analysis, but what I did should get you part of the way there. From where I leave off, importing the data into SAS and then transforming the data from long to wide ( http://www.ats.ucla.edu/stat/sas/modules/longtowide_data.htm) might be a reasonable approach. Hope this helps. Bridget Here's what I did, with the help of my colleague Dan Nordlund, to attach matched pair identifiers to my dataset (shown here using the example lalonde dataset): > data(lalonde) > rownames(lalonde)<-1:614 > lalonde$id_num<-1:614 > head(lalonde) treat age educ black hispan married nodegree re74 re75 re78 id_num 1 1 37 11 1 0 1 1 0 0 9930.0460 1 2 1 22 9 0 1 0 1 0 0 3595.8940 2 3 1 30 12 1 0 0 0 0 0 24909.4500 3 4 1 27 11 1 0 0 1 0 0 7506.1460 4 5 1 33 8 1 0 0 1 0 0 289.7899 5 6 1 22 9 1 0 0 1 0 0 4056.4940 6 > m.out<-matchit(treat ~ re74 + re75 + educ + black + hispan + age, data=lalonde, method="nearest", ratio=1, + exact= c("age","educ")) > m.out Call: matchit(formula = treat ~ re74 + re75 + educ + black + hispan + age, data = lalonde, method = "nearest", ratio = 1, exact = c("age", "educ")) Sample sizes: Control Treated All 429 185 Matched 118 118 Unmatched 311 67 Discarded 0 0 > > > First, I generate a large dataset of all person-level observations that went into the matching process, with the essential output: id_numbers, treatment identifiers, and weights. > # large datafile > a<-data.frame(lalonde$id_num,m.out$treat,m.out$weights) > colnames(a)<-c("id_num","trt","weights") > head(a) id_num trt weights 1 1 1 1 2 2 1 0 3 3 1 1 4 4 1 0 5 5 1 1 6 6 1 1 > Then, I reconfigure the "match.matrix" information so that it is a data frame with variables in numeric format, identifying treated units and matched control units, and a match number. I drop non-matched treatment units, so the maximum match number here should correspond with the number of matched treatment units. > # datafile of matches > b<-as.data.frame(m.out$match.matrix) > colnames(b)<-c("matched_unit") > b$matched_unit<-as.numeric(as.character(b$matched_unit)) > b$treated_unit<-as.numeric(rownames(b)) > #now delete matches=na > c<-b[!is.na(b$matched_unit),] > c$match_num<-1:dim(c)[1] > head(c) matched_unit treated_unit match_num 1 269 1 1 3 350 3 2 5 222 5 3 6 247 6 4 7 460 7 5 9 515 9 6 > head(m.out$match.matrix) 1 1 "269" 2 NA 3 "350" 4 NA 5 "222" 6 "247" > Last, I attach the match numbers to the large data-set of person observations as follows. NA is the match number for treatment units with no matched controls, and for controls that were not matched to any treatment units. > # attach match number to large datafile > a[c$matched_unit,4]<-c$match_num > a[c$treated_unit,4]<-c$match_num > colnames(a)[4]<-"match_num" > head(a) id_num trt weights match_num 1 1 1 1 1 2 2 1 0 NA 3 3 1 1 2 4 4 1 0 NA 5 5 1 1 3 6 6 1 1 4 > A couple of checks to see that the data reconfiguration is working as planned: > # check match numbers > a[!is.na(a$match_num) & a$match_num==1,] id_num trt weights match_num 1 1 1 1 1 269 269 0 1 1 > a[!is.na(a$match_num) & a$match_num==2,] id_num trt weights match_num 3 3 1 1 2 350 350 0 1 2 > a[!is.na(a$match_num) & a$match_num==3,] id_num trt weights match_num 5 5 1 1 3 222 222 0 1 3 > table(a$match_num) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 106 107 108 109 110 111 112 113 114 115 116 117 118 2 2 2 2 2 2 2 2 2 2 2 2 2 2016-09-21 18:00 GMT+02:00 <matchit-request(a)lists.gking.harvard.edu>: > Send Matchit mailing list submissions to > matchit(a)lists.gking.harvard.edu > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.gking.harvard.edu/mailman/listinfo/matchit > or, via email, send a message with subject or body 'help' to > matchit-request(a)lists.gking.harvard.edu > > You can reach the person managing the list at > matchit-owner(a)lists.gking.harvard.edu > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Matchit digest..." > > > Today's Topics: > > 1. Does it have an option to save the number of matched pairs??? > (Andreu Ferrero) > 2. Re: Does it have an option to save the number of matched > pairs??? (Gary King) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 21 Sep 2016 14:40:36 +0200 > From: Andreu Ferrero <fromnorden(a)gmail.com> > To: matchit(a)lists.gking.harvard.edu > Subject: [matchit] Does it have an option to save the number of > matched pairs??? > Message-ID: > <CAH2QH=eTLP4UBAwxfCs0EXjYGu1bXMd5EhpOEzVbB3Dsco6KXw@mail. > gmail.com> > Content-Type: text/plain; charset="utf-8" > > Dear Authors, > > Recently I am using quite a lot the R-package that you build. > It is great, but I have a question. > Does it have an option to save the number of matched pairs??? > I was looking into "match.data()" function, but I did not find it... > > I think this point may be important, beacuse one may use "condicional > logistic regression models" after matching's process. > > I hope you have time to answer me. > > Have a good day, > > Andrew > > -- > Andreu Ferrero Gregori >

7 years, 7 months

Does it have an option to save the number of matched pairs???

by Andreu Ferrero

Dear Authors, Recently I am using quite a lot the R-package that you build. It is great, but I have a question. Does it have an option to save the number of matched pairs??? I was looking into "match.data()" function, but I did not find it... I think this point may be important, beacuse one may use "condicional logistic regression models" after matching's process. I hope you have time to answer me. Have a good day, Andrew -- Andreu Ferrero Gregori

7 years, 7 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Matchit September 2016