```{tip}
Take a look at our [documentation](https://nsquaredlab.github.io/DocOctopy/README.html).
```
## What is this?
DocOctopy , the **research** library for kinematics, kinetics, and everything else you can think of that has to do with EMG and AIs.
> **Important**
> Be aware that this project is used for **research**. Do not expect the same stability as from numpy for example.
## How to install?
> **WARNING**
> Order matters!
- clone
- install [poetry](https://python-poetry.org/docs/#installation)
- based on your hardware:
- GPU (NVIDIA): `poetry install`
- CPU: run `pip install torch torchvision --upgrade` after `poetry install`
- for building the documentation add `docs` to the *with* flag: `poetry install --with docs`
## What is what?
This projects uses the following structure:
- doc_octopy: This is the main package. It contains:
- datasets: Contains data loaders and creators as well as a lot of filters to preprocess the data.
- models: Contains all models and their components.
- utils: Various utilities from training loggers to constants.
- docs: Contains the documentation.
- examples: Contains examples on how to use the package.
- tests: Contains tests for the package.
## What papers/preprints use this package?
