This software is made available for download via this website subject to the license terms and conditions, available also on this website. You will be required to view and accept these license terms and conditions upon installation of the software. If you do not accept or agree to these license terms, you are not permitted to download, install or otherwise use the software in any way. By continuing with this download, you are agreeing to the applicable license terms and conditions.
It is recommended to use at least a dual processor system with 4GB of system memory for GPI.
GPI is built and distributed on PyPI, the official third-party software repository for Python packages distributed using the package manager pip. GPI also uses conda to create it's development enviromet. Conda allows you to create and maintain many separate "environments", each containing different software (such as different versions of Python). You can also save snapshots of an environment to help others verify or reproduce your work using exactly the same software versions.
Note If you do not want terminal windows to open with the "base" conda environment active by default, you can disable this behavior with the following:
conda config --set auto_activate_base false
Launch a new terminal instance. Windows users with Anaconda can launch "Anaconda Prompt" from the Start Menu, or any other terminal (cmd.exe, PowerShell, or MinGW) for which conda has been configured. Next, choose a name for the environment that will hold GPI. We will use
gpi_envfor this tutorial. Finally, run the following commands to set up and configure the new environment:
On Mac & Linux:
conda create -n gpi_env python=3.9 fftw eigen
conda create -n gpi_env python=3.9 fftw eigen pthreads-win32 –c defaults –c conda-forge
Then, activate conda environment:
conda activate gpi_env
Python 3.9 currently provides the "smoothest" running GPI, but you can also install with Python 3.7 or 3.8.
Install GPI and the core nodes. In the same terminal as above (making sure
gpi_envis still the active environment), run the following command:
pip install gpilab
Initialize & compile gpi PyFI nodes
Running GPI - GPI is now installed and ready to use. Run it with the command
gpi. When starting from a new terminal, remember that you will first need to activate the environment
conda activate gpi_env.
You can update GPI like any other pip package, using the command line.
pip install gpilab --upgrade
- If you are having issues with node libraries not being visible, make sure to
~/.gpircfile for the correct library path. By default, the packaged
gpi_corelibrary and any library under