[R] phyloclim help

[David Winsemius]

On Jun 13, 2012, at 11:00 PM, Brent Hendrixson wrote:

>
> Hi David -
>
> Thanks for the response.
>
>> On Jun 13, 2012, at 6:29 PM, Brent Hendrixson wrote:
>>
>>>
>>> Hi all –
>>>
>>> My apologies if this is a redundant message.  I am a newbie to R and
>>> have had a terrible time trying to figure out how to set up an
>>> analysis.  I have installed and loaded the appropriate package
>>> (phyloclim), but I'm not sure I have defined the arguments  
>>> correctly.
>>>
>>> The usage for the command I’m trying to perform is:
>>>
>>> niche.equivalency.test(spec, n, maxent, mx = 2000)
>>
>> You might give the name of the package. (I'm not likely to have any
>> experience with that package, but it might ring a bell so someone  
>> with
>> the right background.)
>
> The package is "phyloclim" (evolutionary biology).
>
>>
>>>
>>> I am comfortable with the arguments "spec", "n", and "mx" but have
>>> had problems with "maxent"
>>>
>>> maxent         is a list containing the location of the MAXENT
>>> application and its input files (see below)
>>>
>>>> From what I have gathered, the maxent argument is a list of four
>>>> elements:
>>>
>>> app                       the path to the MAXENT application (MAXENT
>>> is a JAVA program)
>>>
>>> samples               the path to a SWD-formatted file with sample
>>> points (e.g., samples.csv)
>>>
>>> background         the path to a SWD-formatted file with background
>>> points (e.g., background.csv)
>>>
>>> projections       the path to a folder containing environmental GIS
>>> layers to be used for projection of the MAXENT models
>>>
>>> Suppose all of these files are placed into the directory “C:
>>> \Documents and Settings\hendrb\Desktop\R”.   Can anyone determine  
>>> if
>>> I'm on the right track with the following syntax or am I way off?
>>>
>>>> app <- file.path("C:\\Documents and Settings\\hendrb\\Desktop\\R\
>>>> \maxent.jar")
>>>> samples <- file.path("C:\\Documents and Settings\\hendrb\\Desktop\ 
>>>> \R
>>>> \\allspecies_swd.csv")
>>>> background <- file.path("C:\\Documents and Settings\\hendrb\ 
>>>> \Desktop
>>>> \\R\\background_swd.csv")
>>>> projections <- file.path("C:\\Documents and Settings\\hendrb\
>>>> \Desktop\\R")
>>>> maxent <- list(app, samples, background, projections)
>>>
>>> When I try to execute the command niche.equivalency.test based on
>>> the above information, I get the following statement:
>>>
>>>> niche.equivalency.test(spec, n = 10, maxent, mx = 2000)
>>> Error in read.table(file = file, header = header, sep = sep, quote =
>>> quote,  :
>>>  'file' must be a character string or connection
>>
>> Right. So what is the object named "spec"? The R interpreter is
>> expecting a file name or a connection object.
>
> The object "spec" is a vector of mode = character and length = 2 (it  
> contains the names of two different species that I'm comparing).  I  
> did define it correctly (to my knowledge).  It looks like:
>
> spec <- c("aphonopelma_mojave", "aphonopelma_newmojave")

The error message made clear that someplace along the line there was a  
missing argument to the read.table function.


>
> I'm just unsure if everything else for the "maxent" argument is  
> defined correctly and whether I have set up the list appropriately.   
> If the syntax is okay, then I just need to check the files.   
> Unfortunately, the error message does not indicate which file is  
> throwing stuff off...

Looking at the package documentation I see this:

maxent:  A list containing the location of the MAXENT application and  
its input files (see
details).

Maybe I was wrong about 'spec' and the malformed argument is 'maxent'.  
If I had been constructing it I would have created a _named_ list:

maxent <- list( app =app,
              samples =samples,
              background,   # cannot tell if a csv file meets the  
requiremtn
#                           "The path to a SWD-formatted file with  
background points."
              projections =projections)  # that value .../R  looks  
pretty suspicious, too.

-- 
David

>
> Thanks again!
>
>
>
>>>
>>> I'm incredibly confused by all of this and sincerely appreciate any
>>> help you can provide.
>>>
>>> Regards,
>>>
>>> Brent
>>> 	
>>
>>
>> David Winsemius, MD
>> West Hartford, CT
>>
> 		 	   		

David Winsemius, MD
West Hartford, CT