anamabo / HOBIT-1

Python library for fitting the harmonic oscillator using Bayes optimization and polynomic regression.

HOBIT: Harmonic Oscillator hyBrid fIT

Efficient fit of sine(cosine) functions using a hybrid method

HOBIT is a Python library that combines the power of Hyperopt (https://github.com/hyperopt/hyperopt) with the flexibility of Sklearn oriented to the teaching of physics that is able to fit in a very efficient way functions with the shape

f(x) = y_0 + y_1 * Sin(omega * x + phi)
f(x) = y_0 + y_1 * Cos(omega * x + phi)

whose are commonly used to describe harmonic oscillators.

Install

Requirements

Make sure you have installed:

• pandas
• numpy
• hyperopt
• sklearn

To install HOBIT, go to the directory where the code is located and type in a console:

python setup.py install

Get started

In the folder notebooks you will find four jupyther notebooks with the following examples:

1. Fit of a Cosine function using scipy and description of the gradient descent methond.
2. Fit of a Cosine function using Hyperopt package.
3. Fit of a Cosine function using HOBIT package.
4. Fit of a Cosine function using HOBIT package.

Detailed explanation on the usage of these codes are inside the notebooks.