|
|
![[parent]](http://images.planetmath.org:8080/images/uparrow.png) Viewing Message
|
|
|
| ``Re: "ln" in pstricks''
by stevecheng on 2009-08-01 16:44:53 |
|
| > Steve, you are fully right -- thank you very much!
> I still wonder why the system doesn't allow to continue the
> logsin curve to the interval [Pi/2, Pi].
Sorry, no idea.
> P. S. -- How about your German project?
Unfortunately, I haven't got time. Incidentally the two other software developers in my group both speak German, and one is a Finn (PhD from Courant Institute); Jussi --- you are from Finland too right?
Something that might be interesting to PM'ers: Recently I've been working on optimized computations of transcendental functions, like exp, ln, normal cdf inverse, for use in Monte Carlo simulations (to value financial instruments). We found the implementations of the functions in system libraries were too slow, and because of our environment, the non-open-source replacements from CPU vendors were unsuitable. Knowing nothing about the subject, I had to dig up some research papers, and made notes. Hopefully, with permission, I will be able to publish them.
// Steve
|
| | [ reply | up | top ] | |
|
|
|
|
|
|
|
