Re: mpz_cmp_d and NaNs in = and <

[Marius Vollmer]

Kevin Ryde <address@hidden> writes:

> address@hidden writes:
> >
> > Divide by it and you'll see  :-)
> 
> Oh, well, all I mean is that it's hard to see what it could represent.
> A sort of remnant parity indicating how many positives and negatives
> went into a product or something.  But I don't pretend to any great
> expertise in this sort of area, so maybe it's good for something.

There ought to be some good stuff in

       Branch Cuts for Complex Elementary Functions,
    or Much Ado About Nothing's Sign Bit

    in The State of the Art in Numerical Analysis,
       (eds. Iserles and Powell), Clarendon Press, Oxford, 1987.

but I couldn't find an online copy.

-- 
GPG: D5D4E405 - 2F9B BCCC 8527 692A 04E3  331E FAF8 226A D5D4 E405