problem in 'tolerant' in latent class

[asghar13]

Hi, hope you are well,
I ran ''lc_model = LatentClassSimpleMNL(n_latent_classes=3, fit_method="mle", optimizer="lbfgs", epochs=1000, tolerance=1e-20)
hist, results = lc_model.fit(elec_dataset, verbose=1)''
in your sample in https://colab.research.google.com/github/artefactory/choice-learn/blob/main/notebooks/models/latent_class_model.ipynb#scrollTo=F57NWfxAYp9b
but I got the error of
''

TypeError Traceback (most recent call last)
in <cell line: 2>()
1 lc_model = LatentClassSimpleMNL(n_latent_classes=3, fit_method="mle", optimizer="lbfgs", epochs=1000, tolerance=1e-20)
----> 2 hist, results = lc_model.fit(elec_dataset, verbose=1)

3 frames
/usr/local/lib/python3.10/dist-packages/choice_learn/models/simple_mnl.py in init(self, add_exit_choice, intercept, optimizer, lr, **kwargs)
37 Learning Rate to be used with optimizer.
38 """
---> 39 super().init(add_exit_choice=add_exit_choice, optimizer=optimizer, lr=lr, **kwargs)
40 self.instantiated = False
41 self.intercept = intercept

TypeError: ChoiceModel.init() got an unexpected keyword argument 'tolerance'
''

I know the part of '' ---> 39 super().init(add_exit_choice=add_exit_choice, optimizer=optimizer, lr=lr, **kwargs)
40 self.instantiated = False
41 self.intercept = intercept''
is related to the SimpleMNL, but I got the similar problem for my own dataset also.
I would be grateful for any help to solve the problem

[VincentAuriau]

Hello @asghar13 !

Thank you for trying choice-learn and reaching out.
The argument "tolerance" has been renamed into "lbfgs_tolerance" and it hasn't been well cascaded onto the Latent Class model. Thank you for pointing out this bug.

The issue should be fixed in the branch fix-tol-lcmodel.
I will run a few tests before merging into main and pushing on PyPI to update the pip package. Are you fine working with a clone of the repository in the mean time ?

[asghar13]

Thank you for your time and addressing the issue. I really appreciate your efforts in fixing it. I can wait until the changes are merged into the main branch and the package is updated on PyPI.

In the meantime, I will try with the fix-tol-lcmodel branch as suggested. Thank you again for your support.

[VincentAuriau]

Hello @asghar13,
I have release the v1.1.1 that should fix the issue.
It is also avaialble on PyPI.
Let me know if you encounter any other problem. Thanks again for reporting it ! :)

[asghar13]

thanks a lot