Why IEEE NaN and Inf?

Thread View: sci.math.num-analysis

1 messages

[T] Back to Threads [V] Tree View [G] Group Articles

1 total messages Started by Dan Frankowski Thu, 06 Aug 1992 14:40

#199

Author: Dan Frankowski
Date: Thu, 06 Aug 1992 14:40

21 lines
1358 bytes

I know the behavior of Inf and NaN (1/0 = Inf, 0/0 = NaN, etc.) in the
IEEE floating point arithmetic standard.  In fact, I have the standard
sitting here in front of me.  But what are the reasons they were
implemented?  Why Infs?  NaNs?  Signalling NaNs vs.  quiet NaNs?

This is part of my work to understand the effects of adding a complex
type to C.  For instance, in general

  (x, y) * (u, v) = ( x*u - y*v, y*u + x*v )

So 3.0 * (Inf, 5.0) = (3.0, 0.0) * (Inf, 5.0)
                    = (3.0*Inf - 0.0*5.0, 0.0*Inf + 3.0*5.0)
                    = (Inf, NaN)

But there is a reasonable argument to preserve a scaling property of
real numbers and have 3 * (Inf, 5.0) = (Inf, 15.0).

What would be reasonable behaviors of Inf and NaN to preserve with
complex arithmetic?

Dan Frankowski         dfrankow@centi.cs.umn.edu

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads