-
I suppose currently the categorical distribution only supports probability vector but not logit.
This is a little bit inconvenient in that softmax is also not implemented yet, as such it could be h…
-
### Problem description
All [string functions in the `str` namespace](https://pola-rs.github.io/polars/py-polars/html/reference/series/string.html) should be able to work extremely fast on categorica…
-
Hi, thanks for this great library!
I am getting perfect correlation for the following toy example. I expected ~zero correlation as in the regression approach.
```python
import numpy as np
import…
-
Categorical variables must be scaled by `sqrt(0.5)` per the docs in `kbal`. This fix would be as simple as passing in `mixed_data=TRUE` and `cat_columns=...` for categorical variables. This could ea…
-
Is it throwing an error for you? just wondering I am trying to run processed_data but it isn't working lol
-
# Summary
As a user, I ....
-
Add Example, Discussion
-
For the GROUP BY do not list any categoric variables with more than 10 categories.
![Image](https://github.com/user-attachments/assets/edc95bfc-6f19-4ad4-920a-6067adf3a0d3)
-
When creating an area chart from a polars dataframe with a categorical datatype a ValueError is thrown. This works fine if using `mark_line` instead of `mark_area` or if one converts the polars df to …
-
Hi Kassem,
I like this project.
I recommend replacing XGBoost with CatBoost. CatBoost handles categorical features more efficiently and often requires less hyperparameter tuning. For more infor…