Which algorithms of PyOD support categorical data?

[asmaier]

Which algorithms of PyOD support categorical or mixed continuous and categorical data? I couldn't find any information about that or a code example on how to deal with categorical data.

I also happened to notice that in the benchmark paper (https://arxiv.org/abs/2206.09426) for local, global and dependency anomalies proximity based algorithms like KNN and LOF take the lead. However these algorithms only work on continuous data, not categorical data. Are these benchmark results therefor only valid for datasets with continuous features, where one can define a distance metric?