Author: Emily Quinn Finney
This is my exploration of recommender systems while attending the Recurse Center (https://www.recurse.com/). The original code was written with the intent of being tested on toy databases (charities_toydata.txt and cat_food.txt). In the future, it will be implemented using data obtained from Charity Navigator (https://www.charitynavigator.org/).
The package includes:
- api.py: Code I used to access the Charity Navigator server using its API and Python's requests package.
- charities_realdeal.ipynb (not yet implemented): a Jupyter notebook containing a brief analysis of data obtained from Charity Navigator.
- charities_toyscript.py: an implementation of a content-based recommender system, primarily using toy databases.
- test_api.py: a pytest module that tests the api.py module.
The package also includes a directory of tests (tests/) written for the
charities_toyscript.py module. I used this project as a means of exploring
several test frameworks in Python, so the test directory contains tests written
for Python's unittest, pytest, doctest, and hypothesis libraries.
To run the unittest module, ensure pytest is installed. Then type
pytest -v -s tests/test_unittest.py
To run the pytest module, ensure pytest is installed. Then type
pytest -v -s tests/test_pytest.py
Running the code suffices to run the doctests, and examples are shown in each function's docstrings.
To run the hypothesis module, ensure pytest and hypothesis are installed. Then type
pytest -v -s tests/test_hypothesis.py
Last edited 04/05/2018