[Bioc-devel] Questions about some checks in the latest BiocCheck

First, the issue with using class() is that the return is a vector possibly
with >1 element. This _does_ happen.  This is why is() etc is much better,.

Second, why do you need case insensitivity when testing for class. That
makes NO sense to me. You should be testing for a specific (collection) of
class(es).  Not just anything with a random name.


On Wed, Nov 28, 2018 at 10:44 AM Vincent Carey <stvjc at channing.harvard.edu>
wrote:

Can you provide the repo for your package in development?  These BiocCheck
issues are important and
should be resolved at your end.  We can give some advice on how to get
better results with BiocCheck.

On Wed, Nov 28, 2018 at 10:34 AM Casper Peters <cp100u at hotmail.com> wrote:

So in the package I'm developing, I have a few specific use cases that
make BiocCheck not so happy.



1:


```

* WARNING: Avoid class()== or class()!= ; use is() or !is()

```

So my first problem arises with checking the classes of objects. I want
the interface to be such that I can allow for case insensitivity and/or
allow the object to be NULL in some specific cases. This is easier with
using `class()` explicitly (especially in the case of case sensitivity).



2:


```

 * ERROR: At least 80% of man pages documenting exported objects
      must have runnable examples. The following pages do not:
```

The second problem arises because of some of my exported functions. I

made

parser+download functions for the Fantom5 data, as this was not available
anywhere inside a package yet. Because of the download being big/taking

too

long, I want to avoid running the checks in R CMD check. For this I added
\dontrun{} at the examples, resulting in BiocCheck not liking this.


Other than that I just have some small notes that I still need to take
care of, but with these I am not sure what to do.

Thanks in advance,

-Casper

        [[alternative HTML version deleted]]

--
The information in this e-mail is intended only for th...{{dropped:10}}