you can associate IEEE floats!
-
-
Replying to @FioraAeterna @pikhq
Under what conditions? Clearly (2*x)*0.5 is not same as 2*(x*0.5) for x=DBL_MAX.
1 reply 0 retweets 0 likes -
-
Replying to @FioraAeterna @pikhq
I understood the problem of optimal rearrangement in terms of equivalence...
2 replies 0 retweets 0 likes -
Replying to @RichFelker @pikhq
i work on graphics, so "equivalence" means "within some reasonable epsilon on our test shaders"
1 reply 0 retweets 5 likes -
i'm part of the cabal that tells people to never use == on floats
2 replies 0 retweets 4 likes -
That cabal is known as People Who Undetstand Floating-Point
1 reply 0 retweets 0 likes -
Replying to @science_dot @FioraAeterna and
Aaaaaah, no, my eyes! There are times when it's perfectly appropriate to use exact equality!
2 replies 0 retweets 4 likes -
Replying to @stephentyrone @science_dot and
and 90% of them involve 0, 1, inf, or -inf.... ; - )
1 reply 0 retweets 1 like -
Replying to @FioraAeterna @science_dot and
Of course. The trouble is when you tell people "never use ==", they then look at the LAPACK codebase and say "these guys don't get it!"
3 replies 0 retweets 1 like
Better rule: if you see someone using == they're either a lot stupider or a lot smarter than you. ;-)
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.