Installation

Some common issues in installation have been written up in the FAQ.

You can install PyARPES in an editable configuration so that you can edit it to your needs (recommended) or as a standalone package from a package manager. In the latter case, you should put any custom code in a separate module which you import together with PyARPES to serve your particular analysis needs.

Installation from Source

Using an installation from source is the best option if you want to frequently change the source of PyARPES as you work. You can use code available either from the main repository at GitLab or the GitHub mirror.

  1. Install Miniconda or Anaconda according to the directions

  2. Clone or otherwise download the respository

git clone https://gitlab.com/lanzara-group/python-arpes

  1. Make a conda environment according to our provided specification

cd path/to/python-arpes
conda env create -f environment.yml

  1. Activate the environment

conda activate arpes

  1. Install PyARPES in an editable configuration

pip install -e .

  1. Recommended: Configure IPython kernel according to the Barebones Kernel Installation below

From Package Managers

It is highly recommended that you install PyARPES through conda rather than pip. You will also need to specify conda-forge as a channel in order to pick up a few dependencies. Make sure you don’t add conda-forge with higher priority than the Anaconda channel, as this might cause issues with installing BLAS into your environment. We recommend

conda config --append channels conda-forge
conda install -c arpes arpes

Additional Suggested Steps

  1. Install and configure standard tools like Jupyter or Jupyter Lab. Notes on installing and configuring Jupyter based installations can be found in jupyter.md

  2. Explore the documentation and example notebooks at the documentation site.

Barebones kernel installation

If you already have Jupyter and just need to register your environment. You can do

pip install ipykernel
python -m ipykernel install --user

You can also give the kernel a different display name in Juptyer with python -m ipykernel install --user --display-name "My Name Here".