[Changeset] Re: Single Precision versus double precision NA

[David Bateman]

John W. Eaton wrote:
> On  6-Jun-2008, David Bateman wrote:
> 
> | No, not unless R makes a change to support the new NA value. I think the
> | change can be justified for R as well if they ever want to support
> | single precision values, and any other way to handle this issue is going
> | to be ugly (ie wrapping float/double and overloading the assignment
> | operators to trap the NA values, or some such)
> 
> Since NA in Octave exists partlty for compatibility with R, then I
> think it would be best to find a way to do this so that we can still
> exchange data with R.  What about always writing values in the old
> format when writing binary data, and converting them when reading?
> 

Sure we can do that but before we do shouldn't we ask if R will support
single precision in the future, as if they do they'll face the same
issue? Also what file format is typically used to exchange data with R?

D.