To transform a dataframe

Hello,

The following does what you want but the order of columns is different.


reshape(df1, v.names = c("Cat", "D?but", "Fin"),
	idvar = "Mat", timevar = "Cat", direction = "wide")


Hope this helps,

Rui Barradas

Em 29-06-2013 10:26, Arnaud Michel escreveu:

Hello

I would like to transform the dataframe df1 into df2 (ie copy the data
from several lines for a men/women to only one line by individu men/women)

dput(df1)
structure(list(Mat = c(934L, 934L, 934L, 935L, 935L, 936L, 936L,
936L, 936L, 937L, 937L, 937L, 937L), Nom = structure(c(2L, 2L,
2L, 3L, 3L, 4L, 4L, 4L, 4L, 1L, 1L, 1L, 1L), .Label = c("TTTT",
"XXXX", "YYYY", "ZZZZ"), class = "factor"), Sexe = structure(c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "Masculin",
class = "factor"),
     Cat = c(6L, 7L, 8L, 7L, 8L, 6L, 7L, 8L, 9L, 3L, 4L, 6L, 7L
     ), D?but = structure(c(5L, 7L, 2L, 12L, 8L, 3L, 4L, 10L,
     11L, 13L, 1L, 6L, 9L), .Label = c("01/01/1990", "01/01/2011",
     "01/02/1986", "01/02/1990", "01/07/1986", "01/07/1993", "01/07/1994",
     "01/07/1996", "01/10/2003", "01/11/2002", "01/11/2011", "13/01/1986",
     "23/01/1986"), class = "factor"), Fin = structure(c(2L, 9L,
     10L, 3L, 10L, 5L, 6L, 7L, 10L, 8L, 1L, 4L, 10L), .Label =
c("30/06/1993",
     "30/06/1994", "30/06/1996", "30/09/2003", "31/01/1990", "31/10/2002",
     "31/10/2011", "31/12/1989", "31/12/2010", "31/12/4712"), class =
"factor")), .Names = c("Mat",
"Nom", "Sexe", "Cat", "D?but", "Fin"), class = "data.frame", row.names =
c(NA,
-13L))


dput(df2)
structure(list(Mat = 934:937, Nom = structure(c(2L, 3L, 4L, 1L
), .Label = c("TTTT", "XXXX", "YYYY", "ZZZZ"), class = "factor"),
     Sexe = structure(c(1L, 1L, 1L, 1L), .Label = "Masculin", class =
"factor"),
     Cat4 = c(NA, NA, NA, 4L), D?but4 = structure(c(1L, 1L, 1L,
     2L), .Label = c("", "01/01/1990"), class = "factor"), Fin4 =
structure(c(1L,
     1L, 1L, 2L), .Label = c("", "30/06/1993"), class = "factor"),
     Cat5 = c(NA, NA, NA, NA), D?but5 = c(NA, NA, NA, NA), Fin5 = c(NA,
     NA, NA, NA), Cat6 = c(6L, NA, 6L, 6L), D?but6 = structure(c(3L,
     1L, 2L, 4L), .Label = c("", "01/02/1986", "01/07/1986", "01/07/1993"
     ), class = "factor"), Fin6 = structure(c(2L, 1L, 4L, 3L), .Label =
c("",
     "30/06/1994", "30/09/2003", "31/01/1990"), class = "factor"),
     Cat7 = c(7L, 7L, 7L, 7L), D?but7 = structure(c(2L, 4L, 1L,
     3L), .Label = c("01/02/1990", "01/07/1994", "01/10/2003",
     "13/01/1986"), class = "factor"), Fin7 = structure(c(3L,
     1L, 2L, 4L), .Label = c("30/06/1996", "31/10/2002", "31/12/2010",
     "31/12/4712"), class = "factor"), Cat8 = c(8L, 8L, 8L, NA
     ), D?but8 = structure(c(2L, 3L, 4L, 1L), .Label = c("", "01/01/2011",
     "01/07/1996", "01/11/2002"), class = "factor"), Fin8 = structure(c(3L,
     3L, 2L, 1L), .Label = c("", "31/10/2011", "31/12/4712"), class =
"factor"),
     Cat9 = c(NA, NA, 9L, NA), D?but9 = structure(c(1L, 1L, 2L,
     1L), .Label = c("", "01/11/2011"), class = "factor"), Fin9 =
structure(c(1L,
     1L, 2L, 1L), .Label = c("", "31/12/4712"), class = "factor")),
.Names = c("Mat",
"Nom", "Sexe", "Cat4", "D?but4", "Fin4", "Cat5", "D?but5", "Fin5",
"Cat6", "D?but6", "Fin6", "Cat7", "D?but7", "Fin7", "Cat8", "D?but8",
"Fin8", "Cat9", "D?but9", "Fin9"), class = "data.frame", row.names = c(NA,
-4L))


Any idea ?
Thank you

HI,
Not sure why df2 didn't include Cat3, Debut3, Fin3, instead included columns with all missing values (Cat5, Debut5, Fin5).

library(plyr)
library(reshape2)
df1New<-ddply(df1,.(Mat),transform,castCat=paste0("Cat",Cat),castD?but=paste0("D?but",Cat),castFin=paste0("Fin",Cat))
res<- join_all(list(dcast(df1New,Mat+Nom+Sexe~castCat,value.var="Cat"),dcast(df1New,Mat+Nom+Sexe~castD?but,value.var="D?but"), dcast(df1New,Mat+Nom+Sexe~castFin,value.var="Fin")),by=c("Mat","Nom","Sexe"))
?indx<-grepl("Mat|Nom|Sexe",colnames(res))
?res1<-cbind(res[indx],res[!indx][order(as.numeric(gsub("[[:alpha:]]","",colnames(res)[!indx])))])
res1
#? Mat? Nom???? Sexe Cat3???? D?but3?????? Fin3 Cat4???? D?but4?????? Fin4 Cat6
#1 934 XXXX Masculin?? NA?????? <NA>?????? <NA>?? NA?????? <NA>?????? <NA>??? 6
#2 935 YYYY Masculin?? NA?????? <NA>?????? <NA>?? NA?????? <NA>?????? <NA>?? NA
#3 936 ZZZZ Masculin?? NA?????? <NA>?????? <NA>?? NA?????? <NA>?????? <NA>??? 6
#4 937 TTTT Masculin??? 3 23/01/1986 31/12/1989??? 4 01/01/1990 30/06/1993??? 6
?# ??? D?but6?????? Fin6 Cat7???? D?but7?????? Fin7 Cat8???? D?but8?????? Fin8
#1 01/07/1986 30/06/1994??? 7 01/07/1994 31/12/2010??? 8 01/01/2011 31/12/4712
#2?????? <NA>?????? <NA>??? 7 13/01/1986 30/06/1996??? 8 01/07/1996 31/12/4712
#3 01/02/1986 31/01/1990??? 7 01/02/1990 31/10/2002??? 8 01/11/2002 31/10/2011
#4 01/07/1993 30/09/2003??? 7 01/10/2003 31/12/4712?? NA?????? <NA>?????? <NA>
?# Cat9???? D?but9?????? Fin9
#1?? NA?????? <NA>?????? <NA>
#2?? NA?????? <NA>?????? <NA>
#3??? 9 01/11/2011 31/12/4712
#4?? NA?????? <NA>?????? <NA>
A.K.



----- Original Message -----
From: Arnaud Michel <michel.arnaud at cirad.fr>
To: r-help at r-project.org
Cc: 
Sent: Saturday, June 29, 2013 5:26 AM
Subject: [R] To transform a dataframe

Hello

I would like to transform the dataframe df1 into df2 (ie copy the data 
from several lines for a men/women to only one line by individu men/women)

dput(df1)
structure(list(Mat = c(934L, 934L, 934L, 935L, 935L, 936L, 936L,
936L, 936L, 937L, 937L, 937L, 937L), Nom = structure(c(2L, 2L,
2L, 3L, 3L, 4L, 4L, 4L, 4L, 1L, 1L, 1L, 1L), .Label = c("TTTT",
"XXXX", "YYYY", "ZZZZ"), class = "factor"), Sexe = structure(c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "Masculin", 
class = "factor"),
? ?  Cat = c(6L, 7L, 8L, 7L, 8L, 6L, 7L, 8L, 9L, 3L, 4L, 6L, 7L
? ?  ), D?but = structure(c(5L, 7L, 2L, 12L, 8L, 3L, 4L, 10L,
? ?  11L, 13L, 1L, 6L, 9L), .Label = c("01/01/1990", "01/01/2011",
? ?  "01/02/1986", "01/02/1990", "01/07/1986", "01/07/1993", "01/07/1994",
? ?  "01/07/1996", "01/10/2003", "01/11/2002", "01/11/2011", "13/01/1986",
? ?  "23/01/1986"), class = "factor"), Fin = structure(c(2L, 9L,
? ?  10L, 3L, 10L, 5L, 6L, 7L, 10L, 8L, 1L, 4L, 10L), .Label = 
c("30/06/1993",
? ?  "30/06/1994", "30/06/1996", "30/09/2003", "31/01/1990", "31/10/2002",
? ?  "31/10/2011", "31/12/1989", "31/12/2010", "31/12/4712"), class = 
"factor")), .Names = c("Mat",
"Nom", "Sexe", "Cat", "D?but", "Fin"), class = "data.frame", row.names = 
c(NA,
-13L))


dput(df2)
structure(list(Mat = 934:937, Nom = structure(c(2L, 3L, 4L, 1L
), .Label = c("TTTT", "XXXX", "YYYY", "ZZZZ"), class = "factor"),
? ?  Sexe = structure(c(1L, 1L, 1L, 1L), .Label = "Masculin", class = 
"factor"),
? ?  Cat4 = c(NA, NA, NA, 4L), D?but4 = structure(c(1L, 1L, 1L,
? ?  2L), .Label = c("", "01/01/1990"), class = "factor"), Fin4 = 
structure(c(1L,
? ?  1L, 1L, 2L), .Label = c("", "30/06/1993"), class = "factor"),
? ?  Cat5 = c(NA, NA, NA, NA), D?but5 = c(NA, NA, NA, NA), Fin5 = c(NA,
? ?  NA, NA, NA), Cat6 = c(6L, NA, 6L, 6L), D?but6 = structure(c(3L,
? ?  1L, 2L, 4L), .Label = c("", "01/02/1986", "01/07/1986", "01/07/1993"
? ?  ), class = "factor"), Fin6 = structure(c(2L, 1L, 4L, 3L), .Label = 
c("",
? ?  "30/06/1994", "30/09/2003", "31/01/1990"), class = "factor"),
? ?  Cat7 = c(7L, 7L, 7L, 7L), D?but7 = structure(c(2L, 4L, 1L,
? ?  3L), .Label = c("01/02/1990", "01/07/1994", "01/10/2003",
? ?  "13/01/1986"), class = "factor"), Fin7 = structure(c(3L,
? ?  1L, 2L, 4L), .Label = c("30/06/1996", "31/10/2002", "31/12/2010",
? ?  "31/12/4712"), class = "factor"), Cat8 = c(8L, 8L, 8L, NA
? ?  ), D?but8 = structure(c(2L, 3L, 4L, 1L), .Label = c("", "01/01/2011",
? ?  "01/07/1996", "01/11/2002"), class = "factor"), Fin8 = structure(c(3L,
? ?  3L, 2L, 1L), .Label = c("", "31/10/2011", "31/12/4712"), class = 
"factor"),
? ?  Cat9 = c(NA, NA, 9L, NA), D?but9 = structure(c(1L, 1L, 2L,
? ?  1L), .Label = c("", "01/11/2011"), class = "factor"), Fin9 = 
structure(c(1L,
? ?  1L, 2L, 1L), .Label = c("", "31/12/4712"), class = "factor")), 
.Names = c("Mat",
"Nom", "Sexe", "Cat4", "D?but4", "Fin4", "Cat5", "D?but5", "Fin5",
"Cat6", "D?but6", "Fin6", "Cat7", "D?but7", "Fin7", "Cat8", "D?but8",
"Fin8", "Cat9", "D?but9", "Fin9"), class = "data.frame", row.names = c(NA,
-4L))


Any idea ?
Thank you

Michel ARNAUD
Charg? de mission aupr?s du DRH
DGDRD-Drh - TA 174/04
Av Agropolis 34398 Montpellier cedex 5
tel : 04.67.61.75.38
fax : 04.67.61.57.87
port: 06.47.43.55.31

______________________________________________
R-help at 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.