Skip to content

How to Install Yohou

This guide shows you how to install Yohou and its optional extras.

Install Yohou

# Add to an existing project
uv add yohou

# Or install globally
uv tool install yohou
pip install yohou
conda install -c conda-forge yohou
mamba install -c conda-forge yohou

Plotting Extra

For Plotly-based visualization, install the plotting optional extra:

uv add yohou[plotting]
pip install yohou[plotting]

Core Yohou works without plotly installed. The plotting extra adds plotly and plotly-resampler for interactive charts.

Note

plotly-resampler depends on tsdownsample, which does not yet ship Python 3.14 wheels. On Python 3.14, install without resampler support or disable it at runtime:

from yohou.plotting import set_config
set_config(resampler=False)

Verify Installation

import yohou
print(yohou.__version__)

Development Setup

To contribute to Yohou or install from source:

git clone https://github.com/stateful-y/yohou.git
cd yohou
just install  # Installs dev dependencies + pre-commit hooks
git clone https://github.com/stateful-y/yohou.git
cd yohou
uv sync --group dev
uv run pre-commit install

Optional Packages

yohou-optuna adds OptunaSearchCV for Bayesian hyperparameter optimization. yohou-nixtla wraps the Nixtla statsforecast, mlforecast, and neuralforecast model families into Yohou's API. See Extensions for installation and usage details.

Python Version Support

Yohou supports Python 3.11, 3.12, 3.13, and 3.14.

See Also

With Yohou installed, head to Getting Started to load a dataset and fit your first forecaster.