matched samples, dataframe, panel data
Sorry, it included a case with both values for dummy were 1.
lst7<-lapply(lst6,function(x) {lst<-lapply(x,function(y) y[sum(y$dummy)==1,]);lst[lapply(lst,nrow)>0]})
?res<-do.call(rbind,lapply(lst7,function(x) do.call(rbind,x)))
?row.names(res)<-1:nrow(res)
nrow(subset(res,dummy==0))
#[1] 22
?nrow(subset(res,dummy==1))
#[1] 22
A.K.
----- Original Message -----
From: Cecilia Carmo <cecilia.carmo at ua.pt>
To: arun <smartpink111 at yahoo.com>
Cc: R help <r-help at r-project.org>
Sent: Friday, June 7, 2013 12:03 PM
Subject: RE: [R] matched samples, dataframe, panel data
Sorry,
Something is not ok, because when? I do
nrow(subset(res,res$dummy==0))
[1] 22
nrow(subset(res,res$dummy==1))
[1] 24 There the number? o observatios in the two subsets is not equal. Thanks again, Cec?lia
De: arun [smartpink111 at yahoo.com]
Enviado: sexta-feira, 7 de Junho de 2013 16:43
Para: Cecilia Carmo
Cc: R help
Assunto: Re: [R] matched samples, dataframe, panel data
Hi,
Not sure if this is what you wanted.
res<-do.call(rbind,lapply(lst6,function(x) do.call(rbind,x)))
row.names(res)<-1:nrow(res)
# this combines the list of lists to a data.frame
res[1:4,]
#? firm year industry dummy dimension
#1? ? 1 2000? ? ? 20? ? 0? ? ? 2120
#2? ? 5 2000? ? ? 20? ? 1? ? ? 2189
#3? ? 4 2000? ? ? 20? ? 0? ? ? 3178
#4? ? 7 2000? ? ? 20? ? 1? ? ? 3245
#or
res<-do.call(rbind,lapply(lst6,function(x) do.call(rbind,x)))
res$group<-gsub("(.*\\..*)\\..*$","\\1",rownames(res))
row.names(res)<-1:nrow(res)
res[1:4,]
#? firm year industry dummy dimension? group
#1? ? 1 2000? ? ? 20? ? 0? ? ? 2120 2000.20? #1 group
#2? ? 5 2000? ? ? 20? ? 1? ? ? 2189 2000.20? #1
#3? ? 4 2000? ? ? 20? ? 0? ? ? 3178 2000.20? #2
#4? ? 7 2000? ? ? 20? ? 1? ? ? 3245 2000.20? #2
A.K.
----- Original Message -----
From: Cecilia Carmo <cecilia.carmo at ua.pt>
To: arun <smartpink111 at yahoo.com>
Cc:
Sent: Friday, June 7, 2013 11:33 AM
Subject: RE: [R] matched samples, dataframe, panel data
Thank you very much.
Just a little thing: how can I put it like a dataframe?
Thanks,
Cec?lia
________________________________________
De: arun [smartpink111 at yahoo.com]
Enviado: sexta-feira, 7 de Junho de 2013 16:27
Para: Cecilia Carmo
Assunto: Re: [R] matched samples, dataframe, panel data
Hi,
There could be easier ways...? I am a bit busy now to try other ways.
----- Original Message -----
From: arun <smartpink111 at yahoo.com>
To: Cecilia Carmo <cecilia.carmo at ua.pt>
Cc: R help <r-help at r-project.org>
Sent: Friday, June 7, 2013 11:25 AM
Subject: Re: [R] matched samples, dataframe, panel data
Hi,
May be this helps:
lst1<-split(final3,list(final3$year,final3$industry))
lst2<-lst1[lapply(lst1,nrow)>0]
lst3<-lapply(lst2,function(x) lapply(x$dimension,function(y) x[(y< (x$dimension+x$dimension*0.1)) & (y> (x$dimension-x$dimension*0.1)),]))
lst4<-lapply(lst3,function(x) x[lapply(x,nrow)==2])
lst5<-lapply(lst4,function(x)x[!duplicated(x)])
lst6<-lst5[lapply(lst5,length)>0]
names(lst6)
# [1] "2000.20" "2001.20" "2002.20" "2003.20" "2004.20" "2001.30" "2002.30"
#[8] "2001.40" "2002.40" "2003.40" "2004.40"
lst6["2000.20"]
#$`2000.20`
#$`2000.20`[[1]]
#? firm year industry dummy dimension
#1? ? 1 2000? ? ? 20? ? 0? ? ? 2120
#21? ? 5 2000? ? ? 20? ? 1? ? ? 2189
#
#$`2000.20`[[2]]
#? firm year industry dummy dimension
#16? ? 4 2000? ? ? 20? ? 0? ? ? 3178
#31? ? 7 2000? ? ? 20? ? 1? ? ? 3245
#
#$`2000.20`[[3]]
#? firm year industry dummy dimension
#11? ? 3 2000? ? ? 20? ? 1? ? ? 4532
#6? ? 2 2000? ? ? 20? ? 0? ? ? 4890
A.K.
________________________________
From: Cecilia Carmo <cecilia.carmo at ua.pt>
To: "r-help at r-project.org" <r-help at r-project.org>
Cc: "smartpink111 at yahoo.com" <smartpink111 at yahoo.com>
Sent: Friday, June 7, 2013 9:56 AM
Subject: Re: [R] matched samples, dataframe, panel data
Again my problem, better explained.
#I have a data panel of thousands of firms, by year and industry and
#one dummy variable that identifies one kind of firms (1 if the firm have an auditor; 0 if not)
#and another variable the represents the firm dimension (total assets in thousand of euros)
#I need to create two separated samples with the same number os firms where
#one firm in the first have a corresponding firm in the second with the same
#year, industry and dimension (the dimension doesn't need to be exatly the
#same, it could vary in an interval of +/- 10%, for example)
#My reproducible example
firm1<-sort(rep(1:10,5),decreasing=F)
year1<-rep(2000:2004,10)
industry1<-rep(20,50)
dummy1<-c(0,0,1,1,0,0,1,1,0,1,1,1,0,0,0,0,0,0,1,1,1,1,0,0,0,0,0,0,0,0,1,0,1,0,1,1,1,1,1,0,0,1,0,0,0,0,0,1,1,1)
dimension1<-c(2120,345,2341,5678,10900,4890,2789,3412,9500,8765,4532,6593,12900,123,2345,3178,2678,6666,647,23789,
2189,4289,8543,637,23456,781,35489,2345,5754,8976,3245,1234,25,1200,2345,2765,389,23456,2367,3892,5438,37824,
23,2897,3456,7690,6022,3678,9431,2890)
data1<-data.frame(firm1,year1,industry1,dummy1,dimension1)
data1
colnames(data1)<-c("firm","year","industry","dummy","dimension")
firm2<-sort(rep(11:15,3),decreasing=F)
year2<-rep(2001:2003,5)
industry2<-rep(30,15)
dummy2<-c(0,0,0,0,0,0,1,1,1,1,1,1,1,0,1)
dimension2<-c(12456,781,32489,2345,5754,8976,3245,2120,345,2341,5678,10900,12900,123,2345)
data2<-data.frame(firm2,year2,industry2,dummy2,dimension2)
data2
colnames(data2)<-c("firm","year","industry","dummy","dimension")
firm3<-sort(rep(16:20,4),decreasing=F)
year3<-rep(2001:2004,5)
industry3<-rep(40,20)
dummy3<-c(0,0,1,0,1,0,1,0,1,1,1,1,1,0,0,0,0,1,0,0)
dimension3<-c(23456,1181,32489,2345,6754,8976,3245,1234,1288,1200,2345,2765,389,23456,2367,3892,6438,24824,
23,2897)
data3<-data.frame(firm3,year3,industry3,dummy3,dimension3)
data3
colnames(data3)<-c("firm","year","industry","dummy","dimension")
final1<-rbind(data1,data2)
final2<-rbind(final1,data3)
final2
final3<-final2[order(final2$year,final2$industry,final2$dimension),]
final3
#So my data is final3 is like this:
? firm year industry dummy dimension
26? ? 6 2000? ? ? 20? ? 0? ? ? 781
1? ? 1 2000? ? ? 20? ? 0? ? ? 2120
21? ? 5 2000? ? ? 20? ? 1? ? ? 2189
36? ? 8 2000? ? ? 20? ? 1? ? ? 2765
16? ? 4 2000? ? ? 20? ? 0? ? ? 3178
31? ? 7 2000? ? ? 20? ? 1? ? ? 3245
11? ? 3 2000? ? ? 20? ? 1? ? ? 4532
6? ? 2 2000? ? ? 20? ? 0? ? ? 4890
41? ? 9 2000? ? ? 20? ? 0? ? ? 5438
46? 10 2000? ? ? 20? ? 0? ? ? 7690
2? ? 1 2001? ? ? 20? ? 0? ? ? 345
37? ? 8 2001? ? ? 20? ? 1? ? ? 389
32? ? 7 2001? ? ? 20? ? 0? ? ? 1234
17? ? 4 2001? ? ? 20? ? 0? ? ? 2678
7? ? 2 2001? ? ? 20? ? 1? ? ? 2789
22? ? 5 2001? ? ? 20? ? 1? ? ? 4289
47? 10 2001? ? ? 20? ? 0? ? ? 6022
12? ? 3 2001? ? ? 20? ? 1? ? ? 6593
27? ? 6 2001? ? ? 20? ? 0? ? 35489
42? ? 9 2001? ? ? 20? ? 1? ? 37824
60? 14 2001? ? ? 30? ? 1? ? ? 2341
54? 12 2001? ? ? 30? ? 0? ? ? 2345
57? 13 2001? ? ? 30? ? 1? ? ? 3245
51? 11 2001? ? ? 30? ? 0? ? 12456
63? 15 2001? ? ? 30? ? 1? ? 12900
78? 19 2001? ? ? 40? ? 1? ? ? 389
74? 18 2001? ? ? 40? ? 1? ? ? 1288
82? 20 2001? ? ? 40? ? 0? ? ? 6438
70? 17 2001? ? ? 40? ? 1? ? ? 6754
66? 16 2001? ? ? 40? ? 0? ? 23456
43? ? 9 2002? ? ? 20? ? 0? ? ? ? 23
33? ? 7 2002? ? ? 20? ? 1? ? ? ? 25
3? ? 1 2002? ? ? 20? ? 1? ? ? 2341
28? ? 6 2002? ? ? 20? ? 0? ? ? 2345
8? ? 2 2002? ? ? 20? ? 1? ? ? 3412
48? 10 2002? ? ? 20? ? 1? ? ? 3678
18? ? 4 2002? ? ? 20? ? 0? ? ? 6666
23? ? 5 2002? ? ? 20? ? 0? ? ? 8543
13? ? 3 2002? ? ? 20? ? 0? ? 12900
38? ? 8 2002? ? ? 20? ? 1? ? 23456
64? 15 2002? ? ? 30? ? 0? ? ? 123
52? 11 2002? ? ? 30? ? 0? ? ? 781
58? 13 2002? ? ? 30? ? 1? ? ? 2120
61? 14 2002? ? ? 30? ? 1? ? ? 5678
55? 12 2002? ? ? 30? ? 0? ? ? 5754
67? 16 2002? ? ? 40? ? 0? ? ? 1181
75? 18 2002? ? ? 40? ? 1? ? ? 1200
71? 17 2002? ? ? 40? ? 0? ? ? 8976
79? 19 2002? ? ? 40? ? 0? ? 23456
83? 20 2002? ? ? 40? ? 1? ? 24824
14? ? 3 2003? ? ? 20? ? 0? ? ? 123
24? ? 5 2003? ? ? 20? ? 0? ? ? 637
19? ? 4 2003? ? ? 20? ? 1? ? ? 647
34? ? 7 2003? ? ? 20? ? 0? ? ? 1200
39? ? 8 2003? ? ? 20? ? 1? ? ? 2367
44? ? 9 2003? ? ? 20? ? 0? ? ? 2897
4? ? 1 2003? ? ? 20? ? 1? ? ? 5678
29? ? 6 2003? ? ? 20? ? 0? ? ? 5754
49? 10 2003? ? ? 20? ? 1? ? ? 9431
9? ? 2 2003? ? ? 20? ? 0? ? ? 9500
59? 13 2003? ? ? 30? ? 1? ? ? 345
65? 15 2003? ? ? 30? ? 1? ? ? 2345
56? 12 2003? ? ? 30? ? 0? ? ? 8976
62? 14 2003? ? ? 30? ? 1? ? 10900
53? 11 2003? ? ? 30? ? 0? ? 32489
84? 20 2003? ? ? 40? ? 0? ? ? ? 23
76? 18 2003? ? ? 40? ? 1? ? ? 2345
80? 19 2003? ? ? 40? ? 0? ? ? 2367
72? 17 2003? ? ? 40? ? 1? ? ? 3245
68? 16 2003? ? ? 40? ? 1? ? 32489
15? ? 3 2004? ? ? 20? ? 0? ? ? 2345
35? ? 7 2004? ? ? 20? ? 1? ? ? 2345
50? 10 2004? ? ? 20? ? 1? ? ? 2890
45? ? 9 2004? ? ? 20? ? 0? ? ? 3456
40? ? 8 2004? ? ? 20? ? 0? ? ? 3892
10? ? 2 2004? ? ? 20? ? 1? ? ? 8765
30? ? 6 2004? ? ? 20? ? 0? ? ? 8976
5? ? 1 2004? ? ? 20? ? 0? ? 10900
25? ? 5 2004? ? ? 20? ? 0? ? 23456
20? ? 4 2004? ? ? 20? ? 1? ? 23789
73? 17 2004? ? ? 40? ? 0? ? ? 1234
69? 16 2004? ? ? 40? ? 0? ? ? 2345
77? 18 2004? ? ? 40? ? 1? ? ? 2765
85? 20 2004? ? ? 40? ? 0? ? ? 2897
81? 19 2004? ? ? 40? ? 0? ? ? 3892
I want to keep couples of firms one with dummy=1 and other with dummy=0 that matchs in industry, firm and dimension.
But dimension doesn't need to be exactly the same, it is why I refer an interval of + or - 10%.
For example firm 1 matchs with firm 5, because they have the same year, industry, dimension (10% x 2120 = 212 and 2189-2120<212)
and firm 1 is dummy=0 and firm 5 is dummy=1.
So I want to delete firm 6 because it doesn't macth with any firm, and keep firm 1 and 5.
? ? firm year industry dummy dimension
26? ? 6 2000? ? ? 20? ? 0? ? ? 781
1? ? 1 2000? ? ? 20? ? 0? ? ? 2120
21? ? 5 2000? ? ? 20? ? 1? ? ? 2189
Next,
Now I can match firm 4 with firm 7 and delete firm 8.
36? ? 8 2000? ? ? 20? ? 1? ? ? 2765
16? ? 4 2000? ? ? 20? ? 0? ? ? 3178
31? ? 7 2000? ? ? 20? ? 1? ? ? 3245
And so on...
At the end I want to keep only pairs of firms, matched by year, industry and dimension.
If I separate firms with dummy=1 from firms with dummy=0 in two separated dataframes, I have two matched samples
with the same number of observations. That's what I want.
Thank you,
Cec?lia Carmo
Universidade de Aveiro - Portugal