Clash between `pydocstyle` errors and `property` documentation about atomec HOT 2 CLOSED

Ok thanks for clarifying that and see PR #16.

For consistency I would prefer to go with non-imperative docstrings if you agree.

from atomec.

On this imperative mood thing with regards to properties there is an ongoing discussion: PyCQA/pydocstyle#301

If you follow strictly the rules defined in https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_numpy.html then there is no clash. Change (please note capitalised The ...)

-"""dict of floats: the xc energy, containing keys `x`, `c` and `xc`
-for exchange, correlation and exchange + correlation respectively"""
+""":obj:`dict` of :obj:`float`: The xc energy.
+
+Contains keys `x`, `c` and `xc` for exchange, correlation and
+exchange + correlation respectively.
+"""

But Numpy itself uses imperative mood style, e.g. here 🤷‍♂️

Important is consistency.

from atomec.