Could you please create unit test for this?

Maybe you can refer sklearn-test for details.

Hi @Christakou - thanks for the submission! A few general comments:

1. I notice your code relies on the pandas library. In order to be consistent with the rest of the codebase, you should modify the model to expect numpy arrays instead - no pandas allowed :)
2. In order to demonstrate your code is sound, please include unit tests against an existing implementation of the model. sklearn's DiscriminantAnalysis module looks like it might be relevant here.
3. Please include brief documentation in the NumPy style for the public methods in your model.
4. Style: as much as possible, try to follow pep8 style conventions (e.g., camel_case for method, function, and variable names) to ensure readability. You can autoformat your code with the online Black formatter

You can see the general contribution guidelines here. Let me know if you have questions !