Create and Manage CHIME pipeline environments

This package provides the mkchimeenv command to manage CHIME pipeline environments. On cedar this is probably preinstalled into the chime/python module. If you want to install it yourself (e.g. on your own machine) it can be done using pip in the usual manner:

$ pip install "mkchimeenv @ git+https://github.com/chime-experiment/mkchimeenv.git"

which will install its dependencies and the mkchimeenv command into the Python environment. To use it to create a new CHIME pipeline installation, try doing:

$ mkchimeenv create mychimeenv

It will:

• Create a virtual environment in mychimeenv/venv
• Clone all the CHIME pipeline packages into mychimeenv/code
• Try to determine and then install all their dependencies into the virtual environment
• Perform editable installs of all the CHIME pipeline packages so you can hack on them to your hearts content.
• Download the skyfield data required by caput/ch_util for ephemeris calculations.

To activate the environment just run source mychimeenv/venv/bin/activate.

By default, the repositories are cloned from Github via ssh. If you're a CHIME Collaboration member, make sure you can clone the repositories without requiring a passphrase for the key, either by using a passphrase-less key, or (recommended) using an ssh-agent to keep the unlocked key in memory. You will also need to be added to the chime-experiment organisation, ask a friendly CHIME sysadmin to do this if you haven't already got it setup. To test this works try cloning a small private repository, e.g.

$ git clone ssh://git@github.com/chime-experiment/chimedb_config

If you are not a CHIME Collaboration member, the command will still work for you if you use the --non-chime-member option. This option will omit CHIME-internal packages from the virtual environment, and also clone the repositories using https instead of ssh, which does not require a passphrase-less key to be set up.

When installing on your own computer, the --ignore-system-packages will ignore packages that are already installed into your base Python environment, and will instead perform fresh installs of the required packages into the new virtual environment. This can sometimes be useful for avoiding conflicts with your existing Python setup.

To speed up the creation you can use the --fast option to the create command. This will turn off build isolation when pip is installing the packages, which will give a large speed boost (especially on cedar), but may be less robust.