Getting started

The package can be installed using uv or pip. We recommend the former as it provides faster installation times and better dependency management.

To install uv, follow the instructions at uv’s documentation.

Installation

You can install the package either using uv (recommended) or pip. Editable mode is available for development purposes.

Installation with uv

For a regular installation:

uv add hy2dl

For an editable/development installation:

Clone the repository from GitHub:
```
git clone https://github.com/eduardoAcunaEspinoza/Hy2DL.git
cd Hy2DL
```
Alternatively, download the zip file from GitHub and unzip it, then navigate to the repository root.
Sync the environment and install the package:
```
uv sync
```

This will install Hy2DL and all dependencies specified in the pyproject.toml file in editable mode.

Installation with pip

Note: If you decide to use pip, make sure you already have python “>=3.10” installed in your system.

For a regular installation:

pip install hy2dl

For an editable/development installation:

Clone the repository from GitHub (or download the zip file as described above):
```
git clone https://github.com/eduardoAcunaEspinoza/Hy2DL.git
cd Hy2DL
```
Install the package in editable mode:
```
pip install -e .
```

Data

To use Hy2DL, you will need to download and prepare the datasets. These are not included in the package. Instructions on how to download and prepare the datasets can be found in the data folder of the repository.

Running experiments

After installing the package, you can explore and run different experiments.

In the GitHub repository, you will find the folder notebooks, which contains several examples that can serve as a starting point.

Experiment configurations can be specified in two ways:

As .yml files
Directly as a python dictionary

The folder examples provides ready-to-use configurations that can be directly loaded in the notebooks. In addition, the notebook LSTM_Forecast.ipynb demonstrates how to define the configuration as a Python dictionary, offering a hands-on example.