Re: NaNs in ascii

[Mike Miller]

On Sun, 27 Oct 2002, Dirk Eddelbuettel wrote:

> 2.1.37 is fairly fresh we can't really expect everyone to be quite there
> yet.

Also, when you refer to 2.1.37 on your web pages as "unstable," 2.1.36 is
called "testing" and 2.1.35 is called "stable" don't be surprised if
people stick with 2.1.35 even if you don't want them to...

http://packages.debian.org/cgi-bin/search_packages2.pl?keywords=octave2.1&searchon=names&version=all&release=all


> That said, my Debian packages are available on eight different platforms
> (alpha hppa ia64 i386 mips powerpc sparc s390) and only one installation
> command away.

I appreciate your efforts.  I assume this means that these packages
install under Debian Linux on those cpu architectures.  That is good, but
it won't help me because I'm running Solaris, AIX and IRIX on the machines
I'm using right now (mine is Solaris, the others are in our supercomputing
institute -- www.msi.umn.edu).  I got the supercomputer team to install
Octave even though they are running MATLAB.  I want to develop for Octave
and I don't care so much if the code runs on MATLAB.

If you (or anyone else) can suggest the best way to keep my Solaris/sparc
Octave up to date without too much exertion of time, I'm listening.


> Looks like you guessed badly -- it is fixed in 2.1.37. The warnings
> persists, but we may overlook that:

That's good.  For now I'm just using numbers that don't occur elsewhere in
the data and using them as missing-value codes.  I convert those numbers
to NaNs after loading the data.

Thanks very much!

Mike



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------