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,
Cecília Carmo
Universidade de Aveiro - Portugal
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
