R help - May 2010 - Variable variables using R ... e.g., looping over data frames with a numeric separator

Hello,

I have programmed in PHP a lot, and wanted to know if anyone figured out
Variable variables using R.

For example, I have several dataframes of unequal sizes that relate to L
treatments (1, 2, 3, 4, 5,6, L) ... in this case L=7

fData.1
unique.1
fit.nls.1
summary.nls.1
fit.var.1
summary.var.1
.....
fData.2
unique.2
fit.nls.2
summary.nls.2
fit.var.2
summary.var.2
.....
fData.L
unique.L
fit.nls.L
summary.nls.L
fit.var.L
summary.var.L
========
I want to do something like

for(i in 1:L-1)
{
dataStr = gsub(' ','',paste("fData.",i));
dataVar = eval(dataStr);
 ## GOAL is to grab data frame  'fData.1' and do stuff with it, then in
next
loop grab data frame 'fData.2' and do stuff with it


}

#############

in PHP, I would define the string $dataStr = "final.1" and then
$dataVar $$dataStr which is a variable variables use.

Thanks in advance for any help you can offer or suggest.  My current
solution is to write code in PHP that generates lots of R code.  I would
like to do it all in R, so I don't have to rely on another language.


monte

{x:

	[[alternative HTML version deleted]]

Monte Shaffer <monte.shaffer at gmail.com> writes:
> Hello,
>
> I have programmed in PHP a lot, and wanted to know if anyone figured out
> Variable variables using R.
>
> For example, I have several dataframes of unequal sizes that relate to L
> treatments (1, 2, 3, 4, 5,6, L) ... in this case L=7
You should create a list containing 7 data frames, rather than attempting
to identify them with names containing integers. Then you can process
your data frames in a for loop, or with lapply etc, and things should
generally seem much better.

df.list <- list(fData.1, fData.2, fData.3, fData.4, fData.5, fData.6,
fData.7)

Dan
>
> fData.1
> unique.1
> fit.nls.1
> summary.nls.1
> fit.var.1
> summary.var.1
> .....
> fData.2
> unique.2
> fit.nls.2
> summary.nls.2
> fit.var.2
> summary.var.2
> .....
> fData.L
> unique.L
> fit.nls.L
> summary.nls.L
> fit.var.L
> summary.var.L
> ========>
> I want to do something like
>
> for(i in 1:L-1)
> {
> dataStr = gsub(' ','',paste("fData.",i));
> dataVar = eval(dataStr);
>  ## GOAL is to grab data frame  'fData.1' and do stuff with it,
then in next
> loop grab data frame 'fData.2' and do stuff with it
>
>
> }
>
> #############
>
> in PHP, I would define the string $dataStr = "final.1" and then
$dataVar > $$dataStr which is a variable variables use.
>
> Thanks in advance for any help you can offer or suggest.  My current
> solution is to write code in PHP that generates lots of R code.  I would
> like to do it all in R, so I don't have to rely on another language.
>
>
> monte
>
> {x:
>
> 	[[alternative HTML version deleted]]

On 05/17/2010 03:51 PM, Monte Shaffer wrote:
> for(i in 1:L-1)
> {
> dataStr = gsub(' ','',paste("fData.",i));
> dataVar = eval(dataStr);
>   ## GOAL is to grab data frame  'fData.1' and do stuff with it,
then in next
> loop grab data frame 'fData.2' and do stuff with it
>
>
> }
>    
As Dan Davison said, the more standard R way would be to put all your 
data frames in a list, then iterate over the list.

However, if do want to have your data frames in separate variables, and 
then get each data frame in a loop similar to the code fragment above, 
try something like this:

for (i in 1:(L-1)) {
     dataName <- paste("fData.", i, sep="")
     df <- get(dataName)
     ... do something with data frame df ...
}

You can also give additional arguments to get() to tell it where to look 
(pos=,envir=), and whether to look in parent environments 
(inherits=TRUE/FALSE).

-- Tony Plate