skape-gap

skape-gap is a Python module based on time series modeling and machine learning (trees). See the AUTHORS.rst file for a list of contributors.

Installation

This package is not available on PyPI yet. To install it clone git repository and install the package with a symlink :

pip install -e .

Usage

To use prophet (time serie part) and decision tree (machine learning part) modeling.

from skgap import SkapeGap

model = SkapeGap(time_series='prophet',
                 machine_learning='DT',

                yearly_seasonality = True,
                weekly_seasonality = False,
                daily_seasonality = False,
                seasonality_prior_scale = 0.1,

                max_depth_dt=5)

model.fit(date, X_train, y_train)
model.predict(X_test, 10)

Demo

See notebooks directory for a demo of skgap.