Tutorial on simplification?



When I first tried Macsyma (commercial) several years ago, I was
forever impressed by its help system. You type something like "? How
do I solve ..." and you get a list of examples asking "Are you trying
to do ... or ... ....". Often, I would recognize exactly the type of
problem I wanted to do. Selected that example and you see the syntax,
etc... I have no idea what kind, and how much, work would be involved
in creating such a system for Maxima, but it would certainly be a nice
thing to have. 
In the meantime, the docs pointed in the previous two emails (Robert
Dodier and Dr. Fateman) are worth spending time with,
Milan 

* Robert Dodier <robert.dodier@gmail.com> [2005-09-15 15:39:04 -0600]:

> hello daniel,
> 
> > I had no idea until today that the TRIGRAT function
> > existed.  I happened to find it in one of wxMaxima's
> > pull-down menus, and I clicked it to see what it would
> > do.  There are other simplifications that Maxima seems
> > to be able to perform, but figuring out which commands
> > to use and what order to apply them in is daunting to
> > a newbie.
> 
> well, as mentioned elsewhere, it's pretty daunting to long-time users as well.
> 
> i have some ideas here although nothing to specifically address this problem --
> 
> (1) consider http://maxima.sf.net/docs/tutorial/en/minimal-maxima.pdf
> although that is mostly aimed at broader issues.
> 
> (2) it occurs to me that the refererence manual could be improved
> by a category system. what i mean is something like the categories
> shown here: http://en.wikipedia.org/wiki/Ham_sandwich_theorem
> trigrat, for example, would appear in trigonometry and simplification
> categories.
> 
> (3) i have toyed with the notion of a function that would do something
> like this:
> ideas (sqrt(2)*cos(t-%pi/4))  =>  [trigthis, trigthat, trigtheother,
> ..., trigrat]
> presumably the ideas function would apply some heuristics to generate
> a list of relevant functions. i haven't thought this through so i wouldn't
> be surprised if the idea is flawed from the start.
> 
> sorry i can't be more helpful,
> 
> robert dodier
> 
> _______________________________________________
> Maxima mailing list
> Maxima@math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima

--