Contributing to TensorNets

Contributions to TensorNets are welcome! This guide provides instructions for setting up your development environment and running tests.

Development Environment Setup

  1. Fork and Clone the Repository: Start by forking the TensorNets repository on GitHub and cloning your fork locally.

    git clone https://github.com/<your-username>/tensornets.git
cd tensornets

  2. Create a Virtual Environment: It is highly recommended to work within a virtual environment. You can use conda as specified in the .travis.yml file.

    # Create a conda environment for Python 3.6
conda create -n tensornets-dev python=3.6
source activate tensornets-dev

  3. Install Dependencies: Install the project dependencies. The .travis.yml file provides a good reference for the required packages.

    pip install --only-binary=numpy,scipy numpy scipy cython opencv-python==4.1.0.25
pip install pytest==5.4.3 pytest-pep8
pip install tensorflow==1.14.0  # Or another version you want to test against

  4. Install TensorNets in Editable Mode: Install the package in editable mode so that your changes are reflected immediately.

    pip install -e .

Running Tests

TensorNets uses pytest for its test suite. The tests are configured in .travis.yml and pytest.ini.

Code Style (PEP8)

To check for PEP8 compliance, run the following command from the root of the repository:

PYTHONPATH=$PWD:$PYTHONPATH pytest --pep8 -m pep8

The pytest.ini file contains configurations to ignore certain PEP8 rules.

Unit and Integration Tests

To run the main test suite, use the following command:

PYTHONPATH=$PWD:$PYTHONPATH pytest tests/ --verbose --ignore=tests/all*

This command executes the tests in the tests/ directory, excluding the all* files which are likely for comprehensive, long-running evaluations.

Test Matrix

The project is tested against a matrix of Python and TensorFlow versions, as defined in .travis.yml. This ensures broad compatibility. When you submit a pull request, these tests will be run automatically by Travis CI.

Supported Python Versions: 2.7, 3.6

Supported TensorFlow Versions: A wide range from 1.4.0 up to 2.2.0.

Submitting a Pull Request

  1. Create a new branch for your feature or bug fix.
  2. Make your changes and add tests for them.
  3. Ensure all tests pass locally.
  4. Push your branch to your fork and open a pull request against the master branch of the main TensorNets repository.