Skip to main content

Transform-Invariant Non-Negative Matrix Factorization

Project description

Flake8 Linter Pylint Linter Pytest and Coverage Build Documentation Publish to PyPI Open in Streamlit

Logo

Transform-Invariant Non-Negative Matrix Factorization

A comprehensive Python package for Non-Negative Matrix Factorization (NMF) with a focus on learning transform-invariant representations.

The packages supports multiple optimization backends and can be easily extended to handle application-specific types of transforms.

General Introduction

A general introduction to Non-Negative Matrix Factorization and the purpose of this package can be found on the corresponding GitHub Pages.

Installation

For using this package, you will need Python version 3.7 (or higher). The package is available via PyPI.

Installation is easiest using pip:

pip install tnmf

Demos and Examples

The package comes with a streamlit demo and a number of examples that demonstrate the capabilities of the TNMF model. They provide a good starting point for your own experiments.

Online Demo

Without requiring any installation, the demo is accessible via streamlit sharing.

Local Execution

Once the package is installed, the demo and the examples can be conveniently executed locally using the tnmf command:

  • To execute the demo, run tnmf demo.
  • A specific example can be executed by calling tnmf example <example_name>.

To show the list of available examples, type tnmf example --help.

License

Copyright (c) 2021 Merck KGaA, Darmstadt, Germany

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

The full text of the license can be found in the file LICENSE in the repository root directory.

Usage and Citation

If you use this software, please cite us using the information in CITATION.cff.

Contributing

Contributions to the package are always welcome and can be submitted via a pull request. Please note, that you have to agree to the Contributor License Agreement to contribute.

Working with the Code

To checkout the code and set up a working environment with all required Python packages, execute the following commands:

git checkout https://github.com/emdgroup/tnmf.git ./tnmf
cd tmnf
python3 -m virtualenv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt

Now, you should be able to execute the unit tests by calling pytest to verify that the code is running as expected.

Pull Requests

Before creating a pull request, you should always try to ensure that the automated code quality and unit tests do not fail. This section explains how to run them locally to understand and fix potential issues.

Code Style and Quality

Code style and quality are checked using flake8 and pylint. To execute them, change into the repository root directory, run the following commands and inspect their output:

flake8
pylint tnmf

In order for a pull request to be accaptable, no errors may be reported here.

Unit Tests

Automated unit tests reside inside the folder tnmf/tests. They can be executed via pytest by changing into the repository root directory and running

pytest

Debugging potential failures from the command line might be cumbersome. Most Python IDEs, however, also support pytest natively in their debugger. Again, for a pull request to be acceptable, no failures may be reported here.

Code Coverage

Code coverage in the unit tests is measured using coverage. A coverage report can be created locally from the repository root directory via

coverage run
coverage combine
coverage report

This will output a concise table with an overview of python files that are not fully covered with unit tests along with the line numbers of code that has not been executed. A more detailed, interactive report can be created using

coverage html

Then, you can open the file htmlcov/index.html in a web browser of your choice to navigate through code annotated with coverage data. Required overall coverage to is configured in setup.cfg, under the key fail_under in section [coverage:report].

Building the Documentation

To build the documentation locally, change into the doc subdirectory and run make html. Then, the documentation resides at doc\_build\html\index.html.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

tnmf-0.2.0.tar.gz (6.9 MB view details)

Uploaded Source

Built Distribution

tnmf-0.2.0-py3-none-any.whl (6.9 MB view details)

Uploaded Python 3

File details

Details for the file tnmf-0.2.0.tar.gz.

File metadata

  • Download URL: tnmf-0.2.0.tar.gz
  • Upload date:
  • Size: 6.9 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.6.0 importlib_metadata/4.8.2 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.9.8

File hashes

Hashes for tnmf-0.2.0.tar.gz
Algorithm Hash digest
SHA256 0c7a2f7d9d5686e33f4c818310a2edf9a1bd6e7db84988cba98e3f41340cbb2b
MD5 c26440d00306a324a1441d334ba648ce
BLAKE2b-256 85f185c5e22b603be42ef51ace96bb2cf28d5ab6c7ca9ad51d81a2d54f958d6a

See more details on using hashes here.

File details

Details for the file tnmf-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: tnmf-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 6.9 MB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.6.0 importlib_metadata/4.8.2 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.9.8

File hashes

Hashes for tnmf-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 84b7ffcec3be10f49dcfd56f3dc1ea8677f5adee87fcd73f8e6cb448f7f0634d
MD5 9bd780fa35f5401fbd859877ef0d9c0c
BLAKE2b-256 438aafe173448642ad7ac2792136c14d2e04e127098ad44e5f7e4f3c449a23dc

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page