[R] Indexing variables within lapply?

[Uwe Ligges]

Christian Bieli wrote:

> Hello
> I am using R 2.2.0 with Windows XP.
> I've got a five element list object, each element containing two 
> dataframes of equivalent size.
> 
>  > str(mylist)
> List of 1
>  $ data1:List of 2
>   ..$ data1a       :`data.frame':    77 obs. of  63 variables:
>   .. ..$ var1    : num [1:77] 0.41375 0.00056 1.43040 1.43528 0.61730 ...
>   .. ..$ var2    : num [1:77] 1.154 1.686 0.673 0.800 0.760 ...
>   .. ..$ var3    : num [1:77] 1.245 0.575 0.934 0.461 0.757 ...
>     .
>     .
>     .
>   ..$ data1b      :`data.frame':    77 obs. of  63 variables:
>   .. ..$ var1    : num [1:77] 1284.59    7.59    1.16    8.66 2646.38 ...
>   .. ..$ var2    : num [1:77] 1.231 1.129 0.427 1.132 0.692 ...
>   .. ..$ var3    : num [1:77] 1.36 0.68 1.39 2.54 1.39 ...
>     .
>     .
>     .
>  $ data2:List of 2
>   ..$ data2a       :`data.frame':    77 obs. of  63 variables:
>   .. ..$ var1    : num [1:77] 0.41375 0.00056 1.43040 1.43528 0.61730 ...
>   .. ..$ var2    : num [1:77] 1.154 1.686 0.673 0.800 0.760 ...
>   .. ..$ var3    : num [1:77] 1.245 0.575 0.934 0.461 0.757 ...
>     .
>     .
>     .
>   ..$ data2b       :`data.frame':    77 obs. of  63 variables:
>   .. ..$ var1    : num [1:77] 1284.59    7.59    1.16    8.66 2646.38 ...
>   .. ..$ var2    : num [1:77] 1.231 1.129 0.427 1.132 0.692 ...
>   .. ..$ var3    : num [1:77] 1.36 0.68 1.39 2.54 1.39 ...
>      .
>     .
>     .
> 
> My goal is to do calculations with corresponding variables (e.g. 
> correlation between var1 of data1a and var1 of data1b) within each list 
> element.
> I tried to do it with lapply(data, myfunction,...), but I dont know if 

Without having tested:

myfunction <- function(x, var){
	cor(x[[1]][[var]], x[[2]][[var]])
}
lapply(data, myfunction, var="var1")

Uwe Ligges


> theres a way to index the two vars within lapply.
> There is certainly a way to do it by looping, but lapply is much more 
> elegant.
> 
> Thank you for your help.
> Christian
>