[R] Group several variables and apply a function to the group

[Bert Gunter]

... with() is useful here: e.g. in base R, simply  tapply() or ave() with with()

with(df,ave(x, comn,mi, FUN = sd))

-- Bert



On Sun, Dec 4, 2011 at 1:07 PM, John Kane <jrkrideau at yahoo.ca> wrote:
> ?aggregate should do it
>
> aggregate(df$x,list(df$comn, df$mi), sd)
>
> There are other ways of course
>
> Using the reshape2 package
>
> library(reshape2)
> x1 <- melt(df, id=c("comn", "mi"))
> dcast(x1, comn + mi ~ variable, sd)
>
>
>
>
> --- On Sun, 12/4/11, Aurélien PHILIPPOT <aurelien.philippot at gmail.com> wrote:
>
>> From: Aurélien PHILIPPOT <aurelien.philippot at gmail.com>
>> Subject: [R] Group several variables and apply a function to the group
>> To: R-help at r-project.org
>> Received: Sunday, December 4, 2011, 3:32 PM
>> Dear R-experts,
>> I am struggling with the following problem, and I am
>> looking for advice
>> from more experienced R-users: I have a data frame with 2
>> identifying
>> variables (comn and mi), and an output variable (x). comn
>> is a variable for
>> a company and mi is a variable for a month.
>>
>> comn<-c("abc", "abc", "abc", "abc", "abc", "abc", "xyz",
>> "xyz","xyz", "xyz")
>> mi<- c("1", "1","1", "2", "2", "2", "1", "1", "3", "3")
>> x<- c("-0.0031", "0.0009", "-0.007", "0.1929","0.0087",
>> "0.099","-0.089",
>> "0.005", "-0.0078", "0.67" )
>> df<- data.frame(comn=comn, mi=mi, x=x)
>>
>>
>> For each company, within a particular month, I would like
>> to compute the
>> standard deviation of x: for example, for abc, I would like
>> to compute the
>> sd of x for month1 (when mi=1) and for month2 (when mi=2).
>>
>> In other languages (Stata for instance), I would create a
>> grouping variable
>> (group comnn and mi) and then, apply the sd function for
>> each group.
>>
>> However, I don't find an elegant way to do the same in R:
>>
>> I was thinking about the following: I could subset my data
>> frame by mi and
>> create one file per month, and then make a loop and in each
>> file, use a
>> "by" operator for each comn. I am sure it would work, but I
>> feel that it
>> would be like killing an ant with a tank.
>>
>> I was wondering if anyone knew a more straightforward way
>> to implement that
>> kind of operation?
>>
>> Thanks a lot,
>>
>> Best,
>> Aurelien
>>
>>     [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> 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.
>>
>
> ______________________________________________
> 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.



-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm