Si, Ci, Ei, also more pattern matching examples.



Am Samstag, den 13.12.2008, 15:12 +1100 schrieb Billinghurst, David
(RTATECH):
> > From: Dieter Kaiser
> > 
> > The biggest problem with all this code is that I have not 
> > written the documentation. Perhaps I have more time in the 
> > next weeks to do more work on the documentation.
> 
> I have used your comments in expintegral.lisp as the basis of 
> a section in the special functions chapter of the manual.
> 
> I think it would be quite simple to extend the property based
> integration code in sin.lisp:integrallookups to elliptic functions 
> such as jacobi_sn.  The main problem is that these are two argument 
> functions but the integration variable is first arguement.
> 

Thank you very much for your work on the documentation.

Yes, the lookup mechanism has to be further generalized to support
functions with an arbitrary argument list. First I have added some code
to support the exponential_e and incomplete_gamma functions. I have put
comments into the code about this restriction. We could add further
special code for any other function, but I would like a more general
approach.

This generalization and a generalization in sdiffgrad for functions with
a list an argument are also necessary to support the Hypergeometric
functions.

Dieter Kaiser