Pymer4

Fixes
- Fixes issue in `LogisticRegression` API name change

Compatibility Updates

- This version includes a `noarch` build that should be installable on arm-based macOS platforms (e.g. M1, M2, etc)
- This version drops support for Python 3.7 and adds support for 3.9-3.11

Breaking changes

- This version also uses `joblib` for model saving and loading and drops supported hdf5 files previously handled with the `deepdish` library as it is no longer actively maintained. This means that 0.8.1 will **not** be able to load models saved with earlier versions of `pymer4`!

Fixes
- 119
- 122
- 125

This is a minor release that adds supporting for logistic `Lm` models, likelihood ratio tests for `Lmer` models, fixes numerous bugs.

See the full changelog [here](https://eshinjolly.com/pymer4/new)

If you have trouble installing from conda or a pre-built conda package is not available you can install using pip by first creating a new conda environment:

conda create -n pymer4 python=3.8 'r-lmerTest' 'r-emmeans' rpy2 -c conda-forge
conda activate pymer4
pip install -r requirements.txt
pip install .

- Maintenance release that pins `rpy2 >= 3.4.5,< 3.5.1` due to R -> Python dataframe conversion issue on recent rpy2 versions that causes a [recursion error](https://github.com/rpy2/rpy2/issues/866).
- Pending code changes to support `rpy2 >= 3.5.1` are tracked on [this development branch](https://github.com/ejolly/pymer4/tree/dev_rpy2_3.5.1). **Upcoming releases will drop support for** `rpy2 < 3.5.X`
- Clearer error message when making circular predictions using `Lmer` models

- This version is identical to 0.7.6 but supports R >= 4.1
- Installation is also more flexible and includes instructions for using `conda-forge` and optimized libraries (MKL) for Intel CPUs
- Default installations via `conda-forge` will use `openblas` instead of `mkl`

Bug fixes:
- fixes an issue in which a Lmer model fit using categorical predictors would be unable to use .predict or would return fitted values instead of predictions on new data. Thanks to Mario Leaonardo Salinas for discovering this issue

Behind-the-scenes
- All conda packages for this release make use of the [Intel MKL](https://docs.anaconda.com/mkl-optimizations/index.html) libraries which may result in slight estimate differences and fit times. While this was likely already happening before, it has been made explicit in this release, but is subject to change in the future.
- All conda packages also install `R<4.1.1` which has some new functionality regarding namespaces that are not yet compatible with `pymer4`