as.Date.timeDate

[Achim Zeileis]

Yohan:

> Thanks for the message.
>
> As mentioned in the r-sig-finance, timeDate already exports 
> as.Date.timeDate.

Maybe we should have been more precise in our request: It would be great 
if "timeDate" could fully export "as.Date.timeDate" and not only as an S3 
method.

> The problem is really coming from zoo which overwrites the base function 
> as.Date and do not import as.Date methods from other packages.

Yes, true. However, to make the dependency path not longer than it already 
is, it would be nice if you could fully export your method. Then it is 
found by both generics - the one in base R and the one in zoo.

If you're wondering why zoo masks as.Date(). After zoo had long provided 
an as.Date.numeric, base R decided to break that by adding their own that 
requires an additional "origin" argument without default. For several 
year's we worked around this by simply replacing base:::as.Date.numeric in 
the base NAMESPACE when loading zoo. Then, R-core decided to disallow this 
workaround and now we have to ship our own generic to assure correct 
dispatch.

Thanks & best regards,
Z

> Regards,
> Yohan
>
> On 14 oct. 2011, at 20:37, Gabor Grothendieck wrote:
>
>> Changes in R have necessitated a change in zoo which results in any
>> as.Date method in the timeDate package (or any other package that does
>> not export their as.Date methods) not to be dispatched.
>>
>> This can be fixed by either exporting any as.Date method in the
>> NAMESPACE file (or by importing zoo in the NAMESPACE file).
>>
>> See:
>>
>> https://stat.ethz.ch/pipermail/r-sig-finance/2011q4/008751.html
>>
>> --
>> Statistics & Software Consulting
>> GKX Group, GKX Associates Inc.
>> tel: 1-877-GKX-GROUP
>> email: ggrothendieck at gmail.com
>>
>> _______________________________________________
>> Rmetrics-core mailing list
>> Rmetrics-core at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/rmetrics-core
>>
>
> _______________________________________________
> Rmetrics-core mailing list
> Rmetrics-core at r-project.org
> https://stat.ethz.ch/mailman/listinfo/rmetrics-core
>