TypeError: list indices must be integers or slices, not numpy.float64 about dtreeviz HOT 5 OPEN

Indeed @MohammedAlsayed, the error message should be more precise. If you can make a PR for this and display a more relevant error message it would help.

from dtreeviz.

Hi @MohammedAlsayed,

thanks for your feedback and appreciation for the library.
In the case we have to work with a binary classification problem, by default the classes should be converter to 0 and 1 integer values from the client side.

Not sure if we should add this conversion into the library.

Thanks,
Tudor.

from dtreeviz.

Hi @tlapusan,

Thanks for your reply.

I never faced this issue with other libraries like sklearn, or xgboost in binary classification even though my target is type float. I guess this library should be consistent with other libraries as well. It doesn't make sense that it works perfectly in other libraries then crashes here.

However, if there is no desire to add this feature, then the error message should be more precise than what it shows. Since, it took me time to know what is the issue.

from dtreeviz.

@tlapusan I think I need to be added as a collaborator to be able to pull a request. Shall I pull on master or dev branch?

from dtreeviz.

Can you describe a little more what do you mean by collaborator ? anyone should have permission to make the PR.
Yes, into dev branch. thanks.

from dtreeviz.