Getting Started

Environment setup

First create a subdirectory for yourself in the LBG users space:

mkdir /global/cfs/cdirs/lsst/groups/WLSS/LBG/users/$USER

Next setup the lbg-env tool by adding the following to your ~/.bashrc.ext (or ~/.zshrc.ext):

source /global/common/software/m1727/groups/WLSS/LBG/lbg-env/scripts/setup_lbg_env.sh

Then either start a new shell or re-source your ~/.bashrc.ext/~/.zshrc.ext to pick up the change.

You now have access to the lbg-env tool. This tool provides a number of useful abilities, which you can read about by either running lbg-env help or going to the lbg-env docs. For most people, however, typical usage will likely be

lbgu        # go to your LBG user directory
lbg-env     # activate the latest LBG python environment

You may also wish to run lbg-env setup-jupyter to install a Jupyter kernel for the activated environment which you can use in the online portal.

Repository setup

Go to your subdirectory in the LBG users space and clone this repo:

lbgu
git clone git@github.com:LSSTDESC/lbg-pipelines.git

Finally you can install the pre-commit hooks:

pre-commit install

Now you're ready to work!

Running a pipeline

The simplest way to run a pipeline is with run_debug.sh, which requests a short debug allocation automatically and runs the pipeline inside it:

bash scripts/run_debug.sh <run>

where <run> is the name of a file (without the .yaml extension) under configs/runs/. The pipeline(s) to execute are specified by the pipelines: key inside that run config. See Stages, Pipelines, and Runs for more info.

Contributing

To contribute to this project, see the instructions in Contributing