sparse array length is ambiguous; use getnnz() or shape[0]

[CTI2023]

I am working on a multi-label classification task. Most of the feature values are zero. The dataset is a binary vector. When I tried to generate the counterfactual, I got the error like sparse array length is ambiguous; use getnnz() or shape[0]

import numpy as np

e = dice.generate_counterfactuals(data, total_CFs=5, desired_class=1)

e.visualize_as_dataframe(show_only_changes=True)

[gaugup]

@CTI2023 could you share the dataset and code that caused you to run into this issue?

[rachadlakis]

I had the same problem with OneHotEncoder and it successfully worked when using spare_output = False,

categorical_transformer = OneHotEncoder(handle_unknown='ignore', sparse_output=False)

[CloseChoice]

@rachadlakis could you pls share copy-pastable code that reproduces the issue? It's hard to take a look if one first needs to reverse engineer the issue.