- As much as possible, try to follow the existing format of markdown and code.
- Don't forget to run
pylint ./homemade
before submitting pull requests.
- Create new
README.xx-XX.md
file with translation alongside with mainREADME.md
file wherexx-XX
is locale and country/region codes. For exampleen-US
,zh-CN
,zh-TW
,ko-KR
etc. - You may also translate all other sub-folders by creating related
README.xx-XX.md
files in each of them.
- Make your pull requests to be specific and focused. Instead of contributing "several algorithms" all at once contribute them all one by one separately (i.e. one pull request for "Logistic Regression", another one for "K-Means" and so on).
- Every new algorithm must have:
- Source code with comments and readable namings
- Math being explained in README.md along with the code
- Jupyter demo notebook with example of how this new algorithm may be applied