FixBikeNet

The Python package fixbikenet identifies the most important gaps to fill in a city's bicycle network.

The software downloads and pre-processes data from OpenStreetMap, identifies the gaps, saves the results, creates plots and videos. The source code builds on the code from the research paper Automated Detection of Missing Links in Bicycle Networks.

Publication: https://doi.org/10.1111/gean.12324

Installation

The easy way

Important

As of 2026-05-06, the conda-forge installation is not yet working. We will remove this note once it works.

The best way to install FixBikeNet is using conda and the conda-forge channel:

conda install -c conda-forge fixbikenet

Advanced installations

Set up environment

The main step is to set up a virtual environment fbnenv in which to install the package, and then to use or run the environment.

With Pixi

Installation with Pixi is fastest and most stable:

pixi init fbnenv
pixi add --pypi fixbikenet

At this point you can run fixbikenet in the environment, for example as such:

pixi run python examples/mwe.py

Note

The first time you run code with Pixi, it might take a minute longer, as Pixi resolves the environment's dependencies only at this point.

Alternatively, or if you run into issues, clone this repository and create the environment via the environment.yml file:

pixi init --import environment.yml

With mamba/conda/pip

Alternatively to Pixi, use mamba or conda.

Instructions

[!IMPORTANT]
As of 2026-05-06, the conda-forge installation is not yet working. We will remove this note once it works.

mamba create -n fbnenv -c conda-forge fixbikenet
mamba activate fbnenv

Alternatively, or if you run into issues, clone this repository and create the environment via the environment.yml file:

mamba env create --file environment.yml
mamba activate fbnenv
pip install fixbikenet

Run fixbikenet in Jupyter lab

After having set up the environment above, if you wish to run fixbikenet via JupyterLab, follow the

Instructions

With Pixi

Running fixbikenet in Jupter lab with Pixi is straightforward:

pixi run jupyter lab

An instance of Jupyter lab is automatically going to open in your browser after the environment is built.

With mamba/conda

Using mamba/conda, run:

mamba activate fbnenv
ipython kernel install --user --name=fbnenv
mamba deactivate
jupyter lab

Once Jupyter lab opens, switch the kernel (Kernel > Change Kernel > fbnenv)

With pip

Using pip, run:

pip install --user ipykernel
python -m ipykernel install --user --name=fbnenv
jupyter lab

Once Jupyter lab opens, switch the kernel (Kernel > Change Kernel > fbnenv)

Development installation

If you want to develop the project, clone this repository and create the environment via the environment-dev.yml file:

pixi init --import environment-dev.yml

The developemt environment is called fbnenvdev. Make sure to also read our contribution guidelines.

Usage

We provide a minimum working example in two formats:

• Python script (examples/mwe.py)
• Jupyter notebook (examples/mwe.ipynb)

Repository structure

├── fixbikenet             <- Packaged functions and visualizations
├── tests                   <- Tests to execute to ensure functionality
├── .gitignore              <- Files and folders ignored by git
├── .pre-commit-config.yaml <- Pre-commit hooks used
├── README.md
├── environment.yml         <- Environment file to set up the environment using conda/mamba/pixi

Credits

Development of FixBikeNet was supported by the Danish Innovation Fund (Innovationsfonden).