The dataframe contains information in this format:
Type Value
-------------------
catA one
catB two
catA three
My target is to convert the dataframe to this format (Type's values as columns):
catA catB
-----------------
one -
- two
three -
I have been looking into "dummy variables" but it's not what I need. ¿Could someone give me some ideas?
library(reshape2)
df <- data.frame(Type=c("catA","catB","catA"),value=c("one","two","three"))
df
# Type value
# 1 catA one
# 2 catB two
# 3 catA three
dcast(df,value~Type)
# value catA catB
# 1 one one <NA>
# 2 three three <NA>
# 3 two <NA> two
dcast(df,Type~value)
# Type one three two
# 1 catA one three <NA>
# 2 catB <NA> <NA> two
To preserve the order of value
df$Type <- factor(df$Type,c("catA","catB"))
df$value <- factor(df$value,c("one","two","three"))
dcast(df,Type~value)
# Type one two three
# 1 catA one <NA> three
# 2 catB <NA> two <NA>
dcast(df,value~Type)
# value catA catB
# 1 one one <NA>
# 2 two <NA> two
# 3 three three <NA>
You have a long format table and you want to have a wide format. Use dcast() function from reshape2 package for that.
Since you are not working with a rectangular data set, it may make more sense to store the result in a list. This can be accomplished with unstack:
unstack(df, form=Value~Type)
$catA
[1] "one" "three"
$catB
[1] "two"
data
df <- read.table(header=T, text="Type Value
catA one
catB two
catA three")
