[R] Counting observations of a combined factor
David Winsemius
dwinsemius at comcast.net
Sat Sep 19 16:46:22 CEST 2009
On Sep 19, 2009, at 5:39 AM, Sam Player wrote:
> #I have a dataset with two factor. I want to combine those factors
> into a single factor and count the number of data values for each
> new factor. The following gives a comparable dataframe:
>
> a <- rep(c("a", "b"), c(6,6))
> b <- rep(c("c", "d"), c(6,6))
> df <- data.frame(f1=a, f2=b, d=rnorm(12))
> df
>
> # I use the 'interaction' function to combine factors f1 and f2:
>
> df2 <- data.frame(f3=interaction(df[,"f1"], df[,"f2"]), d=df[,"d"])
> df2
>
> # A count of the first data.frame using factor f1 returns the kind
> of results I am looking for:
>
> count <- as.data.frame(table(df$f1))
> count
>
> # Var1 Freq
> #1 a 6
> #2 b 6
>
> # As does a count using factor2:
>
> count2 <- as.data.frame(table(df$f2))
> count2
>
> # Var1 Freq
> #1 a 6
> #2 b 6
>
> # The same procedure on the second dataframe does not treat the
> levels of factor f3 discreetly, instead giving all possible
> combinations of f1 and f2.
We appear to have a different understanding of the term "discrete".
The interaction function produces all possible combinations of factors
and then table() counts the occurrences of such.
>
> count3 <- as.data.frame(table(df2$f3))
> count3
>
> # Var1 Freq
> #1 a.c 6
> #2 b.c 0
> #3 a.d 0
> #4 b.d 6
>
> I need the results to be:
>
> # Var1 Freq
> #1 a 6
> #2 b 6
Puzzled. You already have such. Why would you want the interaction
function to behave differently?
Did you just want to create a label from f1 and f2? That can be
achieved:
> df2 <- df
> df2$f12 <- with( df2, paste(f1,f2,sep=".") )
> df2
f1 f2 d f12
1 a c -0.52902802 a.c
2 a c -1.07351118 a.c
3 a c 0.63463011 a.c
4 a c 0.26857599 a.c
5 a c 1.57677999 a.c
6 a c 1.08645153 a.c
7 b d -0.60400852 b.d
8 b d -0.06611533 b.d
9 b d 1.00787048 b.d
10 b d 1.48289305 b.d
11 b d 0.54658888 b.d
12 b d -0.67630052 b.d
> count3 <- as.data.frame(table(df2$f12))
> count3
Var1 Freq
1 a.c 6
2 b.d 6
>
> # Any suggestions?
>
> --
> Sam Player, B.Sc.(Hons.) B.A.
> Ph.D. Candidate, Faculty of Agriculture, Food & Natural Resources,
> University of Sydney
>
> Email: splayer at usyd.edu.au
>
> Agroecosystems Research Group
> Room 214 J.R.A. McMillan Building A05
> University of Sydney NSW 2006, Australia
>
> Angkor Research Program
> Room 305 Old Teachers College A22
> University of Sydney NSW 2006, Australia
>
> ______________________________________________
> 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.
David Winsemius, MD
Heritage Laboratories
West Hartford, CT
More information about the R-help
mailing list