[R] warnings although R CMD check runs without any problem

[Robert Gentleman]

Hi,

Christian Kohler wrote:
> Uwe Ligges wrote:
>> Can you give examples?
>>
>> Uwe
>>
>>
>> Christian Kohler wrote:
>>   
>>> Dear list members,
>>>
>>> as I am new to R, I would kindly like to ask for your help.
>>>
>>> I checked a package with the 'R CMD check' command and all worked out
>>> well, (only 'OK' status). Loading its vignette afterwards with
>>> 'source("package.Rcheck/inst/doc/vignette.R")' yields a message saying
>>> that several (well, in my case more than 50) warnings still exist and
>>> can be viewed via the 'warnings()' command.
>>>
>>> My question now is: why are there still warnings even though the check
>>> has been passed without any problems? Generally speaking, would it be
>>> possible to submit this package e.g. to the Bioconductor project, even
>>> if these warnings still exist?
>>>
>>>
>>> Thanks in advance,
>>> Christian
>>>
>>>
>>>
>>>     
>> ______________________________________________
>> 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.
>>   
> Hello Uwe,
> 
> as the package is to submit to Bioconductor, I should have posted to the
> BioC-Help instead but initially it was rather a 'R'-specific question to me.
> My apology for that.
> 
> Examples for warning messages are
> 
>> warnings()
>> 1:The exprSet class is deprecated, use ExpressionSet instead
> ...
>> 6: The phenoData class is deprecated, use AnnotatedDataFrame (with
> ExpressionSet) instead

  These tell you that you are using data structures that are no longer 
supported and should not be used. You need to fix them - the appropriate 
replacements are also indicated in the help message.

>> 13: In s[v == x[1]] <- r : Number of elements to be replaced is not a
> multiple of the substitution length.

   This is also serious. It suggests that you have an error in your 
code. Typically if one replaces the elements in one vector
s[v == x[1]], with those in another vector, namely r, the two sets 
should be the same length. In most (but not all) cases where this 
happens there is a bug.

> ....
> 
> In my opinion, these messages solely indicate some 'hints' which do not
> interfere with any functionality within the package. Right?

   No, wrong - the first two tell you that you are badly (at least one 
year) out of date in your use of data structures and the last one tells 
you that you probably have a serious problem with your code.


> As there is no reference to a file, which triggers these messages, I
> have currently no idea where to take a closer look at.

   Well, you did not give the whole output, and in it there really are 
names of files and locations.  You can also use an editor to search for
  1) any instance of the work exprSet or phenoData and replace those 
parts of the code with the new classes

  2) search for s[v == x[1]], and then use some of R's debugging methods 
to determine whether there is a bug or not.

  best wishes
    Robert

> 
> Regards,
> Christian
> 

-- 
Robert Gentleman, PhD
Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M2-B876
PO Box 19024
Seattle, Washington 98109-1024
206-667-7700
rgentlem at fhcrc.org