Skip to content

INSTALL.md

Latest commit

 

History

History
107 lines (72 loc) · 4.5 KB

INSTALL.md

File metadata and controls

107 lines (72 loc) · 4.5 KB

Installation Guide

⚠️ Note: If you migrate to darts version >=0.41.0 from versions <0.41.0, refer to the migration guidelines below

From PyPI

Darts offers a modular installation system with optional dependencies. Choose the installation that fits your needs:

  • Core only (without neural networks, Prophet, LightGBM, CatBoost, XGBoost, StatsForecast): pip install darts
  • Core + PyTorch (for neural network models): pip install "darts[torch]"
  • Core + Prophet, LightGBM, CatBoost, XGBoost, StatsForecast (no neural networks): pip install "darts[notorch]"
  • All available models (except listed below): pip install "darts[all]"

If the PyTorch installation fails on your platform, please follow the official installation guide for PyTorch, then try installing Darts again.

Optional dependencies for specific models

Some models have additional dependencies that are not included in the all installation option. To use them, you need to install the following dependencies separately from PyPI:

Model Dependencies
NeuralForecastModel neuralforecast>=3.0.0
TiRexModel tirex-ts>=1.4.0

From conda-forge

Create a conda environment (e.g., for Python 3.11): (after installing conda):

conda create --name <env-name> python=3.11

Activate the environment

conda activate <env-name>

As some models have relatively heavy dependencies, we provide four conda-forge packages:

  • Core only (without neural networks, Prophet, LightGBM, CatBoost, XGBoost, StatsForecast): conda install -c conda-forge u8darts
  • Core + PyTorch (for neural network models): conda install -c conda-forge -c pytorch u8darts-torch
  • Core + Prophet, LightGBM, CatBoost, XGBoost, StatsForecast (no neural networks): conda install -c conda-forge u8darts-notorch
  • All available models (except listed below): conda install -c conda-forge -c pytorch u8darts-all

Optional dependencies for specific models

Some models have additional dependencies that are not included in the all installation option. To use them, you need to install the following dependencies separately from conda-forge:

Model Dependencies
NeuralForecastModel neuralforecast>=3.0.0
Prophet prophet>=1.1.1

Some models have dependencies not available on conda-forge. To use them, you need to install the following dependencies separately from PyPI:

Model Dependencies
TiRexModel tirex-ts>=1.4.0

Other Information

Enabling GPU support

Darts relies on PyTorch for the neural network models. For GPU support, please follow the instructions to install CUDA in the PyTorch installation guide.

From Docker

We also provide a Docker image with everything set up for you. For this setup to work you need to have a Docker service installed. You can get it at Docker website.

Pull the latest Darts image.

docker pull unit8/darts:latest

To run it in interactive mode:

docker run -it -p 8888:8888 unit8/darts:latest bash

After that, you can also launch a Jupyter lab / notebook session:

jupyter lab --ip 0.0.0.0 --no-browser --allow-root

Then copy and paste the URL provided by the docker container into your browser to access Jupyter notebook.

Important: Darts PyPI Package Changes As of Version 0.41.0

As of Darts version 0.41.0, we have made changes to our PyPI packages:

  • darts: darts now replaces u8darts with all of its installation options (see section above).
  • u8darts: we will stop maintaining the u8darts package in favor of darts. Version 0.41.0 will be the last released version.

We made these changes to simplify the installation and maintenance of Darts.

Migration from Darts versions <0.41.0 to >=0.41.0

No code changes are required - only package installations changes.

For darts users:

# the original `pip install darts` becomes:
pip install "darts[torch]>=0.41.0"

For u8darts users:

# the original `pip install u8darts[option]` becomes:
pip install "darts[option]>=0.41.0"  # or appropriate extras (e.g. darts[all])
⚡