[R] Functional data analysis - problems with smoothing

spencerg spencer.graves at prodsyse.com
Fri Oct 23 17:17:36 CEST 2009


      'isi' has only 14 values, 'age' has 31.  The help page for 
'smooth.basis' says, "argvals: a vector of argument values correspond to 
the observations in array 'y'." 


      I think 'isi' and 'age' should have the same length. 


      I apologfize if the error message was, uh, cryptic. 


      Hope this helps. 
      Spencer Graves


Peter Ehlers wrote:
> Ben,
>
>  I lose my bet! Good thing my opinion is never worth more than $.02.
>
> Looking at the code for smooth.spline() I see that, although 'argvals'
> is the first argument to the function, the error message is generated
> by the second argument. [** Spencer, is this intentional? **]
>
> So, Ben, try str(rlz) or class(rlz) and see whether it conforms to
> one of the structures described in ?smooth.basis.
>
>  -Peter Ehlers
>
> Benjamin Cheah wrote:
>> Hi Peter and Spencer (and everyone else out there),
>>
>> Thank you for your prompt reply to my post re. FDA - FYI - i'm very 
>> new to R and FDA. Currently half way through a biostats masters 
>> degree in Sydney, Australia! Absolutely loving it and I hope to 
>> include R computing on my future CV!
>>
>> Just to clarify things:
>>
>> When I entered what you suggested, this is what I obtained with my 
>> own dataset....
>>> str(isi)
>>  num [1:14] 1 1.5 2 2.5 3 3.5 4 5 7 10 ...
>>
>> Similarly, for Jim Ramsay's dataset (chapter five of Use R book):
>>> str(age)
>>  num [1:31] 1 1.25 1.5 1.75 2 3 4 5 6 7 ...
>>
>> Doesn't this mean that my data object, 'isi' is numeric?
>>
>> Also, I was looking through Jim Ramsay's datasets - I'm fairly sure 
>> that my vector, 'isi' was similarly organised to his vector, 'age' in 
>> chapter 5 of his new Use R FDA book, so I'm fairly certain that the 
>> data is numeric.... but it obviously isn't....
>>
>> Hope this email clarified things - apologies if it hasn't - still 
>> very new to the terminology and overall feel of R.....
>>
>> The R user network is terrific! Thank you for caring!
>>
>> Ben
>>
>>
>>
>>
>> ________________________________
>> From: Peter Ehlers <ehlers at ucalgary.ca>
>>
>> Cc: R-help at r-project.org
>> Sent: Fri, 23 October, 2009 5:07:15 PM
>> Subject: Re: [R] Functional data analysis - problems with smoothing
>>
>> The error message is pretty clear: regardless of what
>> *you* think, R says that 'isi' is not numeric.
>>
>> Are you sure that 'isi' is not a *factor* object?
>> I'm willing to bet that it is.
>>
>> Use str() to check your data.
>>
>> -Peter Ehlers
>>
>> Benjamin Cheah wrote:
>>> Hi all,
>>>
>>> I'm having major issues with smoothing my functional data
>>>
>>> I'm referring to Jim Ramsay's
>>> examples in his books. The following error message keeps appearing,
>>> despite all my data being numeric.... can anyone kindly offer any 
>>> suggestions?
>>>
>>> isi - vector of argument values - i.e. the independent variable of 
>>> the curves
>>> rlz - data array
>>> TMSfdPar - functional data parameter.
>>>
>>>> TMSfdPar = fdPar(TMSbasis, 4, 0.01)
>>>> TMSfdsmooth = smooth.basis(isi, rlz, TMSfdPar)
>>> Error in smooth.basis(isi, rlz, TMSfdPar) : 'argvals' is not numeric.
>>>
>>> I don't understand why the error message keeps popping up. I've 
>>> tried playing around with Jim Ramsay's datasets and I think my data 
>>> is organised in a similar manner to his, so can't understand what's 
>>> going on.... oh, the frustration!
>>>
>>> Thanks in advance,
>>>
>>> Ben
>>> (the amateur R data analyst and statistician.....)
>>>
>>>
>>>
>>>       
>>> __________________________________________________________________________________ 
>>>
>>>
>>>
>>>     [[alternative HTML version deleted]]
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide 
>>> http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>>
>>>
>>
>>
>>
>>       
>> __________________________________________________________________________________ 
>>
>>
>>
>>     [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide 
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>>
>


-- 
Spencer Graves, PE, PhD
President and Chief Operating Officer
Structure Inspection and Monitoring, Inc.
751 Emerson Ct.
San José, CA 95126
ph:  408-655-4567




More information about the R-help mailing list