Re: [R] Handing significance digits

2010-05-28 Thread Joris Meys
Hi Christofer, I don't know what .Net is doing, but for R these globals are dependent on your machine and platform. ?.Machine ?.Platform Don't know if you can actually hack R into believing otherwise. Did you consider the possibility that the underlying algorithms differ between .Net and R? Chee

Re: [R] Handing significance digits

2010-05-28 Thread ONKELINX, Thierry
n: r-help-boun...@r-project.org > [mailto:r-help-boun...@r-project.org] Namens Christofer Bogaso > Verzonden: vrijdag 28 mei 2010 12:51 > Aan: r-h...@stat.math.ethz.ch > Onderwerp: [R] Handing significance digits > > Hi folks, recently I was trying evaluation of some complex

[R] Handing significance digits

2010-05-28 Thread Christofer Bogaso
Hi folks, recently I was trying evaluation of some complex function having exactly same starting values as well as same algorithm in both R and .Net environment. However at the end point I notice that there are some differences in the reported figures from those two applications (as much as 0.10%).