MOTorNOT v1.0.0

MOTorNOT v1.0.0

First tagged release of this fork. Builds on Robert Fasano's original semiclassical MOT scattering-force model and adds everything needed to simulate a full MOT → optical-dipole-trap transfer:

Added

• CPU/GPU backend — every module dispatches through NumPy or CuPy transparently (get_array_module), so the same code runs on CPU or an NVIDIA GPU with no duplicated logic.
• Optical dipole traps (dipole.py) — the AC-Stark potential for a focused Gaussian beam, a retro-reflected 1D lattice, or two crossed beams, with an exact analytic force (no finite differences).
• Multi-level atomic dynamics (levels.py) — F=2 ↔ excited ↔ dark-F=1 rate equations with a repumper, including an exact (non-stiff) solve for the dark-state population.
• Ensemble diagnostics (diagnostics.py) — temperature, velocity/position distributions and other cloud-level observables.
• MOT → dipole recapture (recapture.py) — classical energy-criterion recapture accounting.
• High-level API (api.py) and a backend-aware RK4 integrate() (integration.py) for driving a simulation end to end without touching the individual physics modules.
• A physics-focused README with the equations behind each block.

Install

pip install motornot-1.0.0-py3-none-any.whl
# or, from source:
pip install git+https://github.com/alexlegoshin/MOTorNOT
# optional GPU acceleration:
pip install "cupy-cuda12x[ctk]"

Assets

• motornot-1.0.0-py3-none-any.whl — wheel, install directly with pip.
• motornot-1.0.0.tar.gz — source distribution.