lume-bmad

Tools for using Bmad in LUME

Installation

This package expects bmad and pytao to be provided by conda.

conda create -n lume-bmad -c conda-forge python=3.11 bmad pytao
conda activate lume-bmad
pip install -e .

For development dependencies:

pip install -e ".[dev]"

Quick import check:

python -c "import pytao, lume_bmad; print('ok')"

Implementation example

The test fixtures in tests/ provide a minimal Tao setup you can use directly:

• tests/fodo.init: Tao initialization file.
• tests/fodo.bmad: lattice file loaded by fodo.init.

Because tests/fodo.init references fodo.bmad with a relative path, keep those two files in the same directory.

from pathlib import Path

from pytao import Tao

from lume_bmad.actions import EleScalarVariable
from lume_bmad.model import LUMEBmadModel

# Reuse the fixture Tao setup from tests/
tao_init = Path("tests") / "fodo.init"
tao = Tao(init_file=str(tao_init), noplot=True)

control_variables = [
    EleScalarVariable(
        name="qf:B1_GRADIENT",
        unit="1/m^2",
        element_name="qf",
        property_name="B1_GRADIENT",
    ),
    EleScalarVariable(
        name="qd:B1_GRADIENT",
        unit="1/m^2",
        element_name="qd",
        property_name="B1_GRADIENT",
    ),
]

model = LUMEBmadModel(
    tao=tao,
    action_variables=control_variables,
    dump_locations=["qf", "qd"],
)

# Read and set model variables
print(model.get(["qf:B1_GRADIENT", "qd:B1_GRADIENT"]))
model.set({"qf:B1_GRADIENT": 0.2})

# Enable beam tracking and read a dumped beam distribution
model.set({"track_type": "beam"})
qf_beam = model.get_value("qf_beam")
print(qf_beam.n_particle)

If you want to run this example from outside the repository, copy both fixture files together and pass the path to that copied fodo.init.