FAIRNB

A Repository for FAIR data handling in Jupyter Notebooks. It shows the usage of the DBRepo and Invenio platforms to create a genre prediction SVM and make its results reproducible.

Installation

Install poetry via pip or conda and run

poetry install

Following packages are additionally needed: git-lfs, libsnfile1 (needed for librosa). Install them with your package manager.

To authenticate with the DBRepo and Invenio servers, rename example-config_.yml located in the config folder to config_.yml and enter your credentials in the placeholders. Credentials are not needed for runs done fully locally.

Usage

The main notebook is located in notebooks/main.ipynb. It executes the full workflow of the project. Alternatively, notebooks/standalone.ipynb also contains the full workflow.

Docker

A docker image is also provided for download on invenio, containing a running jupyter instance with all dependencies installed.

Beware this docker image does not have a config file included and will not be able to authenticate with the DBRepo and Invenio servers. Therefore, if one wants to execute the notebooks locally, only_local must be set to true in the execute method.

Roadmap & Project Status

The project has been done in the context of a bachelor thesis, extensions are not planned for now.

Authors and acknowledgment

Lukas Mahler, https://orcid.org/0000-0002-8985-8139

References