Python Time Series Resources

A living document of Python time series resources

Forecasting

Facebook Prophet is a forecasting procedure impomented in R and Python. It set an industry standard for explainable, scalable, and user-friendly forecasting frameworks.

Paper: Link

Code: prophet on GitHub

NeuralProphet is a python library for forecasting time series data based on Neural Network with PyTorch. The library, as the name implies, was inspired by FaceBook’s Prophet library.

Paper: Arxiv Link

Code: neural_prophet on GitHub

Clustering

Agglomerative clustering

Time series K-Means

Kernel K-means

K-means DBA Clustering

K-medoid clustering

Dynamic Time warping (DTW)

Anomaly Detection

Stumpy is a powerful and scalable Python library that efficiently computes something called the matrix profile, which is just an academic way of saying “for every (green) subsequence within your time series, automatically identify its corresponding nearest-neighbor (grey)”

Papers: Stumpy References

Code: Stumpy on GitHub