> Are these "numerical approximation methods" pythonically possible? >
Yes and that's how they are normally found - not necessarily with Python, but by applying computer simulations of the equations. Generally you calculate values in ever decreasing increments until you get enough accuracy. eg you discover a zero crossingh between 3 and 4, then between 3.3 and 3.4 then between 3.36 and 3.37 and so on...
Caveat: You also need to look out for double crossings within a single step change, so don't make the steps too big. And check the number of roots you expect versus the number you get as an error detection scheme.
I tried to follow your tips, and came up with pinchCubic.py for approximating the real roots of cubic equations.
<http://www.rcblue.com/Python/pinchCubic.py>
I'd appreciate any guidance.
Dick Moores [EMAIL PROTECTED]
_______________________________________________ Tutor maillist - [EMAIL PROTECTED] http://mail.python.org/mailman/listinfo/tutor
