search

mne-tools / mne-bids-pipeline

Automatically process entire electrophysiological datasets using MNE-Python.
https://mne.tools/mne-bids-pipeline/
BSD 3-Clause "New" or "Revised" License
138 stars 65 forks source link

Silence Pandas deprecation warnings from seaborn? #971

Open hoechenberger opened 1 month ago

hoechenberger commented 1 month ago

I just ran the group-average step for sensor data, and I received:

[12:26:29] │ ⏳️ sensor/_99_group_average sub-average Adding decoding results.
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.
  data_subset = grouped_data.get_group(pd_key)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/categorical.py:632: FutureWarning: SeriesGroupBy.grouper is deprecated and will be removed in a future version of pandas.
  positions = grouped.grouper.result_index.to_numpy(dtype=float)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.
  data_subset = grouped_data.get_group(pd_key)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/categorical.py:632: FutureWarning: SeriesGroupBy.grouper is deprecated and will be removed in a future version of pandas.
  positions = grouped.grouper.result_index.to_numpy(dtype=float)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.
  data_subset = grouped_data.get_group(pd_key)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/categorical.py:632: FutureWarning: SeriesGroupBy.grouper is deprecated and will be removed in a future version of pandas.
  positions = grouped.grouper.result_index.to_numpy(dtype=float)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.
  data_subset = grouped_data.get_group(pd_key)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.
  data_subset = grouped_data.get_group(pd_key)
/Users/richardhochenberger/mambaforge/envs/TasIn/lib/python3.12/site-packages/seaborn/_base.py:948: FutureWarning: When grouping with a length-1 list-like, you will need to pass a length-1 tuple to get_group in a future version of pandas. Pass `(name,)` instead of `name` to silence this warning.
  data_subset = grouped_data.get_group(pd_key)

I wonder if we could somehow catch these warnings. I believe I'm running the latest seabird release, 0.13.2

larsoner commented 1 month ago

The way I typically try to silence these is by fixing the upstream package and hoping they release :)