Hi:
To condense the output a bit, here's a slight modification that returns, for
each variable
tested, the p-value, confidence limits and odds ratio estimate:
ff<-function(x,y){
m <- fisher.test(table(x,y))
c(m$p.value, m$conf.int, m$estimate)
}
res <- apply(data1[,LETTERS[1:26]], 2, ff, y=data1[,"pLoss"])
rownames(res)[1:3] <- c('p-value', 'lowerCL', 'upperCL')
t(res)
HTH,
Dennis
On Tue, May 11, 2010 at 1:04 AM, Mike White <mikewhite....@btconnect.com>wrote:
> Set up a function for the fisher.test on a 2x2 table and then include this
> in the apply function for columns as in the example below. The result is a
> list with names A to Z
>
> # set up a dummy data set with 100 rows
> Cat<-LETTERS[sample(1:6,100, replace=T)]
> GL<-sample(1:6, 100, replace=T)
> dat<-matrix(sample(c(0,1),100*27, replace=T), nrow=100)
> colnames(dat)<-c(LETTERS[1:26],"pLoss")
> data1<-data.frame(Cat, GL, dat)
>
> # define function fro fisher.test
> ff<-function(x,y){
> fisher.test(table(x,y))
> }
>
> # apply function to columns A to Z
> results<-apply(data1[,LETTERS[1:26]],2, ff, y=data1[,"pLoss"])
> # the results are in the form of a list with names A to Z
> results$C
>
>
> On 19:59, Venkatesh Patel wrote:
>
>> ---------- Forwarded message ----------
>> From: Dr. Venkatesh<drve...@liv.ac.uk>
>> Date: Sun, May 9, 2010 at 4:55 AM
>> Subject: R apply() help -urgent
>> To: r-help@r-project.org
>>
>>
>> I have a file with 4873 rows of 1s or 0s and has 26 alphabets (A-Z) as
>> columns. the 27th column also has 1s and 0s but stands for a different
>> variable (pLoss). columns 1 and 2 are not significant and hence lets
>> ignore
>> them for now.
>>
>> here is how the file looks
>>
>> Cat GL A B C D E F G H I J K L M N O P
>> Q
>> R S T U V W X Y Z pLoss
>> H 5 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> E 5 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> P 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> P 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> F 6 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> E 4 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> H 5 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> J 4 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> J 4 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> E 5 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
>> 0
>> 0 0 0 0 0 0 0 0 0 1
>> S 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
>> 0
>> 0 1 0 0 0 0 0 0 0 1
>> ..
>> ..
>> ..
>> ..
>> ..
>> ..
>>
>> Alphabets A-Z stand for different categories of protein families and pLoss
>> stands for their presence or absence in an animal.
>>
>> I intend to do Fisher's test for 26 individual 2X2 tables constructed from
>> each of these alphabets vs pLoss.
>>
>> For example, here is what I did for alphabet A and then B and then C....
>> so
>> on. (I have attached R-input.csv for your perusal)
>>
>>
>>
>>> data1<- read.table("R_input.csv", header = T)
>>> datatable<- table(data1$A, data1$pLoss) #create a new datatable2 or 3
>>>
>>>
>> with table(data1$B.. or (data1$C.. and so on
>>
>>
>>> datatable
>>>
>>>
>> 0 1
>> 0 31 4821
>> 1 0 21
>>
>> now run the Fisher's test for these datatables one by one for the 26
>> alphabets :(
>>
>> fisher.test(datatable), ... fisher.test(datatable2)...
>>
>> in this case, the task is just for 26 columns.. so I can do it manually.
>>
>> But I would like to do an automated extraction and fisher's test for all
>> the
>> columns.
>>
>> I tried reading the tutorials and trying a few examples. Cant really come
>> up
>> with anything sensible.
>>
>> How can I use apply() in this regard? or is there any other way, a loop
>> may
>> be? to solve this issue.
>>
>> Please help.
>>
>> Thanks a million in advance,
>>
>> Dr Venkatesh Patel
>> School of Biological Sciences
>> University of Liverpool
>> United Kingdom
>>
>>
>>
>>
> ______________________________________________
> R-help@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.
>
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
