FermiBadgerPlugins - Use Badger to run Xopt at Fermilab

Use this repo is one of two ways:

1. Install and run by hand:

git clone git@github.com:fermi-ad/FermiBadgerPlugins.git
cd FermiBadgerPlugins
conda create -n badger-env python=3.12.1 badger-opt=1.4.4 -y
conda activate badger-env 
pip install "acsys[settings]"==0.12.8 --extra-index-url https://www-bd.fnal.gov/pip3 --no-cache-dir
pip install xsuite
# Modify the four "..._ROOT" directories in the config.yaml file.  See notes.
badger -g -cf config.yaml

Nota Bene pip install of acsys-python requires being on the FNAL private network or active VPN thereto.

First-time startup notes:

1. Edit the *_ROOT directories in the config.yaml:

• BADGER_ARCHIVE_ROOT and BADGER_LOGBOOK_ROOT should be some location capacious enough for accumulating some data and logging histories. They can be the same value.
• BADGER_PLUGIN_ROOT and BADGER_TEMPLATE_ROOT can be set to the plugins directory of the cloned repository you are in now.

2. UNCHECK THE Automatic VARIABLES CHECKBOX before doing anything else on the GUI. A bug needs to be addressed.
3. Load the template "TuneQx.yaml" to see a quick-start example using the SimpleVirtualAccelerator Environment. "TuneQxQy_mobo.yaml" adjusts both quad busses to achieve setpoints in both transverse tunes. For both of these setpoints see the Parameters of the environment.

Detailed use guide to the Badger GUI.

On the EAF remember to set export OMP_NUM_THREADS=8 to prevent slow performance from thread flail.

2. or Build a container image for running with Docker or podman

• To build: (may require changes to your Docker descktop for sufficient RAM or other memory resources)

    cd root-of-this-repo/
    docker build -t adregistry.fnal.gov/ml-autotune/testimage:latest --platform linux/amd64 .
    docker login -u $USER adregistry.fnal.gov
    ## Services password ##
    docker push  adregistry.fnal.gov/ml-autotune/testimage:latest  

• To run

    docker run --rm docker run --rm -v .:/app testimage:latest 

• To run your script directly without rebuilding the image:

   docker run --rm -v .:/app testimage:latest