You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It'll be useful to add an is_linear attribute to Parameter so that, if at fitting time all parameters are found to be linear it fits using least squares instead of non-linear least squares. This should dramatically increase the fitting speed in those cases.
TODO
[ ] Add is_linear attribute to all components.
[ ] Add linear least squares to fitters
[ ] Automatically use linear least squares when the problem is found to be linear
[ ] User Guide documentation
[ ] Docstring documentation
[ ] Tests
The text was updated successfully, but these errors were encountered:
I've taken this up again as of this week. The fitting algorithm works, but not yet with convolution. I'd really enjoy it if someone would check out the PR (#1462), where I've merged the linear_parameters PR.
It'll be useful to add an
is_linear
attribute toParameter
so that, if at fitting time all parameters are found to be linear it fits using least squares instead of non-linear least squares. This should dramatically increase the fitting speed in those cases.TODO
[ ] Add
is_linear
attribute to all components.[ ] Add linear least squares to fitters
[ ] Automatically use linear least squares when the problem is found to be linear
[ ] User Guide documentation
[ ] Docstring documentation
[ ] Tests
The text was updated successfully, but these errors were encountered: