validation_curve and learning_curve should support fit_params

[jnothman]

Similar functions like cross_validation and cross_val_score support fit_params to pass sample_weight etc into the model being fitted. These other model selection routines should probably support the same.

[ido-sh]

Hi @jnothman, can I give it a shot?

[jnothman]

Sure, have a go. Remember to include or update a test

[ggc87]

Hi @jnothman,
I was checking this issue and I just saw that the tests for the validation_curve are in test_learning_curve, which is testing the deprecated implementation in learning_curve.py.
Should this test moved in test_cross_validation where validation_curve is implemented now?

[jnothman]

don't worry about the deprecated module. tests should be in model_selection/tests

[ggc87]

Right, I didn't see that the tests are in the model_selection/tests folder, I just saw the test_validation.py in the main test folder.

[gxyd]

@ggc87 @ido-sh Is anyone here working on the issue? I have a pull request ready locally, but since you already asked to work on the issue, I'll wait if you still want to work on it.

[ido-sh]

@gxyd if yours is ready then please feel free to go ahead

[ggc87]

@gxyd same here I was waiting for @ido-sh since he asked before.