[BioC] IRanges causes bug in complete.cases

Kasper Daniel Hansen khansen at stat.berkeley.edu
Tue Dec 15 16:30:42 CET 2009


On Dec 11, 2009, at 17:10 PM, Seth Falcon wrote:

> On 12/10/09 12:17 PM, Kasper Daniel Hansen wrote:
>> To avoid stuff like this in the future, perhaps it would make sense
>> to think of some kind of testing for packages making an S4 generic
>> out of a "base" function.
> 
> Do you have a specific idea of how we could achieve this?  Adding a sanity test, like running the example, for a base function within a package that promotes the function into a generic would have helped this case, but might not catch other issues?

It seems that this issue was caused by some recent changes to the methods package, in which case it is not too strange things break (irritating, but that's life :)

There are now a number of core Bioconductor packages that promotes many functions to generics.  I would suggest seeing if R passes make check with these packages loaded.  For example Biostrings and IRanges.

But yes, I think it would a lot of sense if - in general - R CMD check makes sure that if a package promotes a function from base/utils/... into a generic, that the examples connected to that function still runs.  There may be variants not caught by this check, but I don't think it hurts to have as many unit tests as we can think of.  There may be some implications of this that I have not thought through.

Kasper


More information about the Bioconductor mailing list