Skip to content

Linux/macOS

The installation of CYTools on Linux or macOS is primarily done using conda. A conda environment allows us to package all of the dependencies and ensure that everything will work properly. CYTools can also be installed with pip, but a few optional dependencies (such as Normaliz) are not available that way.

Requirements

  • Linux: Almost any distribution, on a modern x86-64 or arm64 processor.
  • macOS: A recent version of macOS on a Mac with an Apple silicon (M-series) processor. Intel-based Macs are not supported.
  • (Optional but highly recommended) A conda installation. Miniforge is recommended. The installation instructions below assume you are using conda, except for the pip installation.

Installation instructions

  1. Install conda, if not already installed. Installation instructions can be found here.

  2. Run the following command in your terminal.

    curl -fsSL https://cy.tools/install.sh | bash

  3. Enjoy CYTools! 🎉

You can take a look at exactly what the install script does by viewing it at https://cy.tools/install.sh (that's the point of open-source software!). In short, it downloads an environment.yml file and then creates a conda environment from this file.

  1. Install conda, if not already installed. Installation instructions can be found here.

  2. Download the source code of the latest stable release of CYTools from the following link, and extract the zip or tar.gz file into your desired location. Alternatively, clone the repository with git.

    Download

  3. (Optional) Modify environment-dev.yml and any other files to fit your needs.

  4. Navigate to the root directory of CYTools in your terminal, and run conda env create -f environment-dev.yml. This will create a conda environment with CYTools, as well as its dependencies. CYTools is installed in editable mode, so any changes you make to its source code take effect without having to reinstall it.

This method sets up a conda environment (offering to install Miniforge if conda is not found) and adds a CYTools launcher to your applications, which opens JupyterLab in the CYTools environment.

  1. Clone the repository, or download and extract the source code of the latest release.

  2. Navigate to the root directory of CYTools in your terminal, and run the installer for your operating system.

    bash scripts/linux/install.sh  # On Linux
bash scripts/macos/install.sh  # On macOS

To uninstall the app, run scripts/linux/uninstall.sh or scripts/macos/uninstall.sh in the same way.

CYTools and its dependencies can be installed from PyPI with the following command. We recommend doing this in a virtual environment.

pip install cytools

Note that the pip installation does not include Normaliz, which is needed for the computation of Hilbert bases. Use one of the conda installation methods if you need it. The Mosek optimizer and JupyterLab are also not included, but they can be installed with pip install mosek jupyterlab.

Usage

The CYTools environment will be named either cytools if the easy installation or the desktop app was used, or cytools-dev if the advanced installation was followed. To enter this environment, type conda activate cytools or conda activate cytools-dev, respectively. In this environment, cytools can be imported in the same way as any other Python package such as NumPy, SciPy, Matplotlib, etc. JupyterLab is installed by default, and it can be opened with jupyter lab. To exit the conda environment, simply run conda deactivate.

Updating

To update CYTools to the latest version, activate the cytools environment and run pip install --upgrade cytools. If you followed the advanced installation, update the source code instead (for example, with git pull).

Removal

To remove the CYTools environment, run conda env remove --name cytools or conda env remove --name cytools-dev, depending on which one you installed. If you installed CYTools with pip, run pip uninstall cytools instead.

Troubleshooting

Since there are many different platforms, it is possible that our installation scripts don't always work. If this is the case, please let us know by emailing us at [email protected] or by opening an issue on GitHub.