[R] Removing a single element from an array

[David Winsemius]

On Sep 18, 2011, at 11:10 AM, David Winsemius wrote:

>
> On Sep 18, 2011, at 10:47 AM, Dennis Fisher wrote:
>
>> R 2.13.1
>> OS X
>>
>> Colleagues
>>
>> I frequently encounter a situation in which I want to remove a  
>> single element of an array.  For example, if I am reading in a  
>> bunch of CSV files, I create the list of files to be read with:
>> 	LIST	<- dir()
>> However, sometimes I want to exclude one or more files.  I can  
>> accomplish this with a second command:
>> 	LIST	<- LIST[LIST != "filename.to.be.excluded"]
>> This is cumbersome -- is there some more efficient code to  
>> accomplish this?
>
> Doesn't negative indexing provide this?
>
> ?Extract

The most general way to convert character vectors to numbers for this  
purpose is with grep which therefore supports regex patterns:

LIST <- dir(path="~/")
 > length(LIST)
[1] 551
 > LIST[1]
[1] "_train_1.dat"

 > LIST <- LIST[ -grep("_train_1.dat", LIST) ]

 > length(LIST)
[1] 550

Patterns can also be used in the dir() call.

-- 

David Winsemius, MD
West Hartford, CT