pybear-dask 0.2.0

Python modules for data analytics applications with dask

pybear-dask is a Python computing library that supplements the pybear library with analogous modules that have dask capability.

Website: https://github.com/PylarBear/pybear-dask

License

BSD 3-Clause License. See License File.

---

Installation

Dependencies

pybear-dask requires:

• Python (>=3.10, <3.13)

• dask (<2025.1.0)

• dask-ml (<2025.1.0)

• distributed (<2025.1.0)

• pybear (>=0.1.22)

User installation

Install pybear-dask from the online PyPI package repository using pip:

(your-env) $ pip install pybear-dask

Conda distributions are expected to be made available sometime after release to PyPI.

---

Usage

The folder structure of pybear-dask is nearly identical to scikit-learn. This is so those that are familiar with the scikit layout and have experience with writing the associated import statements have an easy transition to pybear-dask. The pybear-dask subfolders are base and model_selection.

You can import pybear-dask’s packages in the same way you would with scikit. Here are a few examples of how you could import and use pybear-dask modules:

from pybear-dask.model_selection import GSTCVDask

search = GSTCVDask()
search.fit(X, y)

from pybear-dask import model_selection as ms

search = ms.AutoGridSearchCVDask()
search.fit(X, y)

---

Major Modules

AutoGridSearchCVDask

Perform multiple uninterrupted passes of grid search with dask_ml GridSearchCV and dask objects utilizing progressively narrower search grids.

• Access via pybear-dask.model_selection.AutoGridSearchCVDask.

GSTCVDask (GridSearchThresholdCV for Dask)

Perform conventional grid search on a classifier with concurrent threshold search using dask objects in parallel and distributed environments. Finds the global optima for the passed parameters and thresholds. Fully compliant with the dask_ml/scikit-learn GridSearchCV API.

• Access via pybear-dask.model_selection.GSTCVDask.

AutoGSTCVDask (AutoGridSearchThresholdCV for Dask)

Perform multiple uninterrupted passes of grid search with pybear-dask GSTCVDask utilizing progressively narrower search grids.

• Access via pybear-dask.model_selection.AutoGSTCVDask.

---

Changelog

See the changelog for a history of notable changes to pybear-dask.

---

Development

Important links

• Official source code repo: https://github.com/PylarBear/pybear-dask

• Download releases: https://pypi.org/project/pybear-dask/

• Issue tracker: https://github.com/PylarBear/pybear-dask/issues

Source code

You can clone the latest source code with the command:

git clone https://github.com/PylarBear/pybear-dask.git

Contributing

pybear-dask is not ready for contributions at this time!

Testing

pybear-dask 0.2 is tested via GitHub Actions to run on Linux, Windows, and MacOS, with Python versions 3.10, 3.11, and 3.12. pybear-dask is not tested on earlier versions, but some features may work.

If you want to test pybear-dask yourself, you will need:

• pytest (>=7.0.0)

The tests are not available in the PyPI pip installation. You can get the tests by downloading the tarball from the pybear-dask project page on pypi.org or cloning the pybear-dask repo from GitHub. Once you have the source files in a local project folder, create a poetry environment for the project and install the test dependencies. After installation, launch the poetry environment shell and you can launch the test suite from the root of your pybear-dask project folder with:

(your-pybear-dask-env) you@your_computer:/path/to/pybear-dask/project$ pytest tests/

Project History

This project was spun off the main pybear project just prior to the first public release of both. pybear-dask was spun off to ensure maximum stability for the main pybear project, while keeping these modules available.

Help and Support

Documentation

Documentation is not expected to be made available via a website for this package. Use the documentation for similar packages in the main pybear package. See the repo for pybear: https://github.com/PylarBear/pybear/

Communication

• GitHub Discussions: https://github.com/PylarBear/pybear-dask/discussions

• Website: https://github.com/PylarBear/pybear-dask