We currently analyze a sample of the dataset which is then used to determine the encoder to use. For the dictionary encoder we should consider penalize it choosing it since it has slower encode/decode compared to other encoding algorithms with comparatively similar bits per val.
We currently analyze a sample of the dataset which is then used to determine the encoder to use. For the dictionary encoder we should consider penalize it choosing it since it has slower encode/decode compared to other encoding algorithms with comparatively similar bits per val.