Install | HierarchicalForecast - Nixtla

Install HierarchicalForecast with pip or conda

We recommend using uv as Python package manager, for which you can find installation instructions here. You can then install the released version of HierarchicalForecast:

uv pip install hierarchicalforecast

Alternatively, you can directly install from the Python package index with:

pip install hierarchicalforecast

or within a conda environment:

conda install -c conda-forge hierarchicalforecast

Tip We recommend installing your libraries inside a python virtual or conda environment.

Installing from source

We recommend using uv as Python package manager, for which you can find installation instructions here.

Clone the HierarchicalForecast repo:

$ git clone https://github.com/Nixtla/hierarchicalforecast.git && cd hierarchicalforecast

Create the environment:

$ uv venv --python 3.10

Activate the environment:

on MacOS / Linux:

$ source .venv/bin/activate

on Windows:

$ .\.venv\Scripts\activate

Install the dependencies and the library

uv pip install -r setup.py
uv pip install .