NeuralProphet#

Fusing traditional time series algorithms using standard deep learning methods, built on PyTorch, inspired by Facebook Prophet and AR-Net.

Simple Example#

>>> from neuralprophet import NeuralProphet
>>> m = NeuralProphet()
>>> metrics = m.fit(df)
>>> forecast = m.predict(df)
>>> m.plot(forecast)

Features#

NeuralProphet provides many time series modeling and workflow features, in a simple package:

  • Support for global modeling of many time series.

  • Automatic selection of training related hyperparameters.

  • Plotting utilities for forecast components, model coefficients and final predictions.

  • Local context through Autoregression and lagged covariates.

  • Changing trends and smooth seasonality at different periods.

  • Modeling of event, holiday, and future regressor effects.

  • Many customization options, such as regularization.

Resources#