Skip to main content

Python project for inverse modeling of accelerator lattices

Project description

Documentation Status PyPI versions PyPI pyversions

(acc)-1

accinv - Python project for inverse modeling of accelerator lattices


Note: This project is in proof-of-concept stage and therefore lacks more advanced features of some of the implemented methods.


In its current state, the package provides functionality for inverse modeling of linear optics via fitting of orbit response matrix (ORM) data, typically referred to as linear optics from closed orbits. It supports cpymad as a backend.

The main class is accinv.loco.Loco which requires one of the models from accinv.model, as well as a method for computing Jacobians, as an argument. Two methods for Jacobian computation are available.

  • AnalyticalJacobianMethod: This method uses an analytical formula to compute the Jacobian of the ORM with respect to quadrupole gradient errors and BPM and steerer gain errors. The data for the analytical formula is obtained from a single Twiss call for the current lattice configuration.
  • NumericalMJacobianMethod: This method uses a finite difference approximation scheme to compute the Jacobian of the ORM with respect to the quadrupole gradient errors. Thus, the number of ORMs that will be computed is proportional to the number of quadrupoles.

The inverse modeling process can be started by creating a Loco object and calling its run method:

from accinv.jacobian import AnalyticalJacobianMethod
from accinv.loco import Loco, OrmMeasurement
from accinv.model import Madx

model = Madx(path='path/to/script.madx')
loco = Loco(
    model_and_jacobian_method=(model, AnalyticalJacobianMethod),
    quadrupoles=[...],  # names of quadrupoles
    hbpms=[...],        # names of horizontal BPMs
    hsteerers=[...],    # names of horizontal steerers
    vbpms=[...],        # names of vertical BPMs
    vsteerers=[...],    # names of vertical steerers
    orm_measurement=OrmMeasurement(
      orm=np.load('path/to/measured_orm.npy'),
      uncertainty=np.load('path/to/orm_uncertainty.npy'),
  ),
)
result = loco.run()

Please consider the documentation of the Loco class for more details.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

accinv-0.1.0.post1.tar.gz (25.0 kB view details)

Uploaded Source

Built Distribution

accinv-0.1.0.post1-py3-none-any.whl (26.9 kB view details)

Uploaded Python 3

File details

Details for the file accinv-0.1.0.post1.tar.gz.

File metadata

  • Download URL: accinv-0.1.0.post1.tar.gz
  • Upload date:
  • Size: 25.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.9.0

File hashes

Hashes for accinv-0.1.0.post1.tar.gz
Algorithm Hash digest
SHA256 70a0c14afe95f0a198a26df4e42991dd7a5e73da62be281899cc18ea22e52653
MD5 490c21867664fd01ac72e6bceddf95a3
BLAKE2b-256 05c45fbaa981a01d7a86b748878c10167320d4787bb0482f75705c227503bab6

See more details on using hashes here.

File details

Details for the file accinv-0.1.0.post1-py3-none-any.whl.

File metadata

File hashes

Hashes for accinv-0.1.0.post1-py3-none-any.whl
Algorithm Hash digest
SHA256 ccc367bad22f8322b469593c4f1ba1ae1759374f9fac9dd0aa5b0996699c1383
MD5 6cf614ebcea2b632a39952b28cfc0980
BLAKE2b-256 a00b63e01bd1462b3a1c5bdcd18523264d5ff85e4a99130795ab2c374d2c6ca3

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page