77 improve matrix class #112
Conversation
… bounds for get_solution
…. Also start using np.allclose for convergence so removed own routines
There was a problem hiding this comment.
Hey Sam,
Some minor comments. I believe we should implement the numpy supported testing routines when we use arrays in tests. Also I would like you to add some more descriptive text near the sparse matrix conversion routines.
A discussion point might be the small formatting changes required when we touch classes. I would suggest to revise the class inititialization when we make small changes. What do you think?
| self.add_equation(equation) | ||
| if dumb: | ||
| self.dump_matrix() | ||
| if len(equations) > self.num_unknowns: |
There was a problem hiding this comment.
Maybe we can move this to another method something like: verify_equations? You need to call it in the solve method but it is a separate routine.
There was a problem hiding this comment.
Hey Sam,
Some minor comments. I believe we should implement the numpy supported testing routines when we use arrays in tests. Also I would like you to add some more descriptive text near the sparse matrix conversion routines.
A discussion point might be the small formatting changes required when we touch classes. I would suggest to revise the class inititialization when we make small changes. What do you think?
There was a problem hiding this comment.
Hey Sam,
Some minor comments. I believe we should implement the numpy supported testing routines when we use arrays in tests. Also I would like you to add some more descriptive text near the sparse matrix conversion routines.
A discussion point might be the small formatting changes required when we touch classes. I would suggest to revise the class inititialization when we make small changes. What do you think?
|
Discussed with Michiel to approve and merge |
Refactored the matrix class to use