Sigmoidal 0.2.0

Sigmoidal is a small library to allow you to fit and evaluate sigmoid functions in a way that works like the Numpy Polynomial class.

Sigmoidal

Sigmoidal is intended to work like the Numpy Polynomial class where it makes sense. Specifically it supports:

• Using the sigmoid function in a very natural way like y = sig(x) including when x is a numpy array. (Once you've created a Sigmoid instance sig = Sigmoid(...))
• Fitting a Sigmoid to data just like Polynomial with Sigmoid.fit(x, y).
• Taking the first or second derivative with deriv().
• Finding the roots of the sigmoid or it's first or second derivitive with .roots().
• The convenience method .linspace() which can get you an array of dependent values with only the range of independent values.
• .copy()
• Operations ==, !=, str(), repr()

Running Tests

• python -m unittest discover tests

Setup for Deployment

• pip install twine

Building the Package

• python setup.py sdist bdist_wheel

Deploying

• twine upload --repository-url https://upload.pypi.org/legacy/ dist/*