Jan Tilly

Actually: With Pandas=1.2.4: ```python import pandas as pd x1 = pd.Series(pd.Categorical(["a"], categories=["a", "b"])) x2 = pd.Series(pd.Categorical(["a"], categories=["b", "a"])) x2.astype(x1.dtype) # > 0 a # > dtype: category # > Categories...

> What does the proposed fixed behavior look like? I would just write ```python X[k] = X[k].astype(dtype) if not X[k].cat.categories.equals(dtype.categories): X[k] = X[k].cat.reorder_categories(dtype.categories) ``` instead of https://github.com/dask/dask-ml/blob/0ea276da1d78db582f40e1c256dfca4f70e6cbc6/dask_ml/preprocessing/data.py#L568

Sorry, I have no idea. I'm not using visual studio. Do you also run into this problem with the original version of inih (https://github.com/benhoyt/inih)? Just googling this for a second,...

You get a set of the sections via ``` std::set sections = reader.Sections() ``` You can convert that into any other data structure that you find convenient. But I'm not...

xref https://github.com/Quantco/tabmat/issues/75

In applications where we need to drop a base level, we typically one-hot encode our categoricals before using glum. With few levels, that's also faster than using categorical types. As...

https://github.com/conda-forge/cffi-feedstock/pull/47 was just merged and `cffi=1.15.1=*_3` is now also available from `conda-forge`. This fixes the issue for me :partying_face: Thank you!

I have another example for this issue: I could pin down a specific version of `sqlalchemy` (_not_ `snowflake-sqlalchemy`) that introduced this. Take this file and call it `t.py`: ```python import...

Adding to the post above: for recent versions of `sqlalchemy`, behavior did in fact change with the 1.4.x releases of `snowflake-sqlalchemy`. Take the Python snippet from above and create the...

1.4.4 fixes the issue for me. Thank you!