On 13-Feb-08 12:40:48, Barry Rowlingson wrote: > hadley wickham wrote: > >> It's more than that as though, as floating point addition is >> no longer guaranteed to be commutative or associative, and >> multiplication does not distribute over addition. Many concepts >> that are clear cut in pure math become fuzzy in floating point >> math - equality, singularity of matrices etc etc. > > I've just noticed that R doesn't calculate e^pi - pi as equal to 20: > > > exp(pi)-pi == 20 > [1] FALSE > > See: http://www.xkcd.com/217/ > > Barry
Barry, These things fluctuate. Once upon a time (sometime in 1915 will do) you could get $[US]4.81 for £1.00 sterling. One of the rare brief periods when the folks on opposite side of the Atlantic saw i^i (to within .Machine$double.eps, which at the time was about 0.001, if you were lucky and didn't make a slip of the pen). R still gets it approximately right: 1/(1i^1i) [1] 4.810477+0i $i^i = £1 Best wishes, Ted. -------------------------------------------------------------------- E-Mail: (Ted Harding) <[EMAIL PROTECTED]> Fax-to-email: +44 (0)870 094 0861 Date: 13-Feb-08 Time: 15:57:02 ------------------------------ XFMail ------------------------------ ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
