Installation#

Installing from PyPi#

ETNA is available on PyPI, so you can use pip to install it.

Install default version:

pip install --upgrade pip
pip install etna

The default version doesn’t contain all the dependencies, because some of them are needed only for specific models, e.g. Prophet, PyTorch. Available user extensions are the following:

  • prophet: adds prophet model,

  • torch: adds models based on neural nets,

  • wandb: adds wandb logger,

  • auto: adds AutoML functionality,

  • statsforecast: adds models from statsforecast,

  • classiciation: adds time series classification functionality,

  • chronos: adds Chronos-like pretrained models,

  • timesfm: adds TimesFM pretrained models.

  • clearml: adds ClearML logger with tensorboard support

Install extension:

pip install "etna[extension-name]"

Install all the extensions:

pip install "etna[all]"

Installing from the repository#

ETNA can also be installed from the repository. It allows you to get the latest version of the library or version from a specific commit. It can be done by a command:

pip install "etna[all]@git+https://github.com/etna-team/etna.git@master"

where you could replace master branch with some other identifier.

Installing on computing platforms#

The library, starting with version 3.0, can be installed on the Google Colab and Kaggle platforms. To do this, you can use any of the instructions above.

On Google Colab you shouldn’t forget to restart the environment after installing the library.

Installing on Apple M1 (ARM)#

We are trying to make ETNA work with Apple M1 and other ARM chips. However due to novelty of these architectures some packages ETNA depends on are going to run into some problems or bugs.

List of known problems:

Possible workaround:

  • Use python>=3.9 and initialize virtualenv.

  • Build CatBoost via instruction in the comment above: you will need llvm installed via brew and you should specify paths to llvm and python binaries in flags correctly to make successful build.

  • Install built CatBoost whl in virtualenv.

  • Install library: LLVM_CONFIG="/opt/homebrew/Cellar/llvm@11/11.1.0_3/bin/llvm-config" pip install etna==<version>. (LLVM_CONFIG flag may be optional and it could be different a little bit in version spec but you should have 11 or 12 major version)