Reorganizing data from 3 rows to 1

问题

I need to reorganize data from a csv file that contains mostly repeating data. I have the data imported into R in a dataframe but I am having trouble with the following:

ID   Language  Author   Keyword
12   eng       Rob      COLOR=Red
12   eng       Rob      SIZE=Large
12   eng       Rob      DD=1
15   eng       John     COLOR=Red
15   eng       John     SIZE=Medium
15   eng       John     DD=2

What I need to do is transform this into a row with each keyword in a separate column

ID   Language  Author  COLOR  SIZE      DD
12   eng       Rob     Red    Large     1

Any ideas?

回答1:

Using the reshape2 package this is straightforward:

With tt defined as in Gary's answer

library("reshape2")

tt <- cbind(tt, colsplit(tt$Keyword, "=", c("Name", "Value")))
tt_new <- dcast(tt, ID + Language + Author ~ Name, value.var="Value")

which gives

> tt_new
  ID Language Author COLOR DD   SIZE
1 12      eng    Rob   Red  1  Large
2 15      eng   John   Red  2 Medium

回答2:

Using plyr ans strsplit you can do something like this :

library(plyr)
res <- ddply(dat,.(ID,Language,Author),function(x){
        unlist(sapply(strsplit(x$Keyword,'='),'[',2))
})

colnames(res)[4:6] <- c('COLOR','SIZE','DD')

 ID Language Author COLOR   SIZE DD
1 12      eng    Rob   Red  Large  1
2 15      eng   John   Red Medium  2

Edit: Here is a generalization that addresses @Brian's concern:

res <- ddply(dat,.(ID,Language,Author), function(x){
             kv <- strsplit(x$Keyword, '=')
             setNames(sapply(kv, `[`, 2),
                      sapply(kv, `[`, 1)) })

回答3:

Try this using reshape2:

tt <- read.table(header=T,text='ID   Language  Author   Keyword
 12   eng       Rob      COLOR=Red
 12   eng       Rob      SIZE=Large
 12   eng       Rob      DD=1
 15   eng       John     COLOR=Red
 15   eng       John     SIZE=Medium
 15   eng       John     DD=2')

tt$Keyword <- as.character(tt$Keyword)

tt <- transform(tt, key_val = lapply(tt$Keyword,function(x) strsplit(x,'=')[[1]][2]),
 key_var = lapply(tt$Keyword,function(x) strsplit(x,'=')[[1]][1]))

tt_new <- dcast (tt, ID + Language + Author ~ key_var, value.var='key_val')

来源：https://stackoverflow.com/questions/15032270/reorganizing-data-from-3-rows-to-1

标签

csv

reshape2