matrix class

[Dan Bishop]

On Jun 12, 7:31 pm, DarrenWeber <Darren.We... at radiology.ucsf.edu>
wrote:
> Below is a module (matrix.py) with a class to implement some basic
> matrix operations on a 2D list.  Some things puzzle me about the best
> way to do this (please don't refer to scipy, numpy and numeric because
> this is a personal programming exercise for me in creating an
> operational class in pure python for some *basic* matrix operations).
>
> 1.  Please take a look at the __init__ function and comment on the
> initialization of the list data with respect to unique memory
> allocation.
>
> 2.  In the operator overloading of __add__, __sub__, etc., the
> statement isinstance(q, Matrix) raises exceptions every time.  This
> statement works fine outside of the class definition, but not during
> the operator evaluation.  What is going here?
...
>
>     def __add__(self, q):
>         'matrix addition: m3 = m1 + m2'
> #         if isinstance(q, Matrix):
> #             arg = ("q is not a matrix instance", q)
> #             raise TypeError, arg

Wouldn't it make more sense to raise an exception if q is NOT an
instance of Matrix?