[SciPy-dev] Linalg2 working and isnan questions?

[Prabhu Ramachandran]

>>>>> "TO" == Travis Oliphant <oliphant.travis at ieee.org> writes:

    TO> I noticed before that some people were saying isnan tests
    TO> failed.

    TO> Could you describe that problem again.

    TO> I got failures on the test

    TO> isnan(log(-1))==1

    TO> as well but only because the scipy.log function returns
    TO> complex values for negative numbers now.  fastumath.log still
    TO> does it the old way for any die-hards.

My original report of this is here:

   http://www.scipy.net/pipermail/scipy-dev/2002-March/000664.html

I havent updated my cvs tree since so its status quo here.

prabhu