R devel - Sep 2007 - to overcome error while computing sd for each subset of data (PR#9931)

Full_Name: Anjali Shitole
Version: R 2.5.1
OS: 
Submission from: (NULL) (202.141.157.91)


I want to compute coefficient of variation for each subset of data.
Each subset is heterogeneous means number of rows in each subset are
different.These all subsets are in one dataset situated one by one.
i have used aggregate command to calculate mean for each subset.
But same command for calculating sd gives error 
see the following program,
> a<-read.table("c:/mri/data/ratioinput.txt",header=T,
na.strings "NA",fill=TRUE)
> dim(a)
[1] 32515    27
> BusinessRisk<-a[,16]/(a[,26]/a[,27]) 
> a1<-a[,1:3]
> a2<-cbind(a1,BusinessRisk)
> c1<-aggregate(BusinessRisk,list(a2[,1]),mean)
> names(a)
 [1] "Co_Code"        "Co_Name"        "Year"     
"TotSholderFund"
 [5] "TotLiab"        "NetBlock"      
"Investments"    "Inventories"
 [9] "SundryDebts"    "CashandBank"   
"TotCurrAssets"  "CurrLiab"
[13] "TotCurrLiab"    "TotAssets"      "NetSales" 
"OpPft"
[17] "Interest"       "GrossPft"       "AdjNetPft"
"Dividend"
[21] "PrefDivi"       "EquityPaidUp"  
"CurrentAssets"  "PBIT"
[25] "TotExpen"       "Equity"         "FaceValue"
> c2<-aggregate(BusinessRisk,list(a2[,1]),sd(BusinessRisk,na.rm=TRUE))
Error in FUN(X[[1L]], ...) : argument "INDEX" is missing, with no
default
> 
Please help me to overcome this error.

anjalishitole wrote:
> 
> Full_Name: Anjali Shitole
> Version: R 2.5.1
> OS: 
> Submission from: (NULL) (202.141.157.91)
> 
>   [snip]
> 
> Please help me to overcome this error.
> 
> 
  Dear Mr. Shitole,

 * this problem is not a bug, please don't post it as such
 * this problem is inappropriate for the r-devel list, which is for
technical questions
and discussion about the workings of R, not for questions about specific
analyses
 * please read the posting guide (URL listed in every message posted),
provide
a reproducible example, and resubmit your question to r-help
 * hint: the third argument to aggregate needs to be a function, further
optional
arguments can be specified.  sd(x) is not a function.  sd is.

 sincerely
    Ben Bolker

-- 
View this message in context:
http://www.nabble.com/to-overcome-error-while-computing-sd-for-each-subset-of-data-%28PR-9931%29-tf4532694.html#a12981381
Sent from the R devel mailing list archive at Nabble.com.