spacopt 0.3.1

spacopt is a package for bringing optimization techniques to spacal-simulation application

spacopt

spacopt - short for spacal-optimization

Description

spacopt is a package for bringing optimization techniques to spacal-simulation application, created for LHCb ECAL studies for different types of calorimeters, such as spacal and shashlik. Should be considered as complementary to spacal-simulation, hosted at gitlab under CERN domain. Up to this point, the main package for optimization is considered Hyperactive.

Features

• Create config files, with user defined parameters of the module.
• Run a MC simulation, using pyton script
• Run Optimization for finding best user-defined parameters of module, to minimize the loss function: $ \dfrac{a}{\sqrt{E}}+b,$ where $a$ - sampling term, $b$ - constant term.

In fact, both $a$ and $b$ could be considered as independent subjects to minimize, as well as other functions of one or both of them.

Installation

pip install spacopt

Usage

import spacopt

# Run Simulation

Contributing

License

• Free software: MIT license
• Documentation: https://spacopt.readthedocs.io.

History

0.3.1 (2022-05-19)

• Readability changges to Simulation prints

0.3.0 (2022-05-19)

• New method to Simulation: MultiFittingMultProc
  • Adds multiprocessing to running geant4 instances (core per every energy point), so now energy points run in parallel.
• Other minor changes to Simulation init: so it now accepts parameters of the module, instead of final method.

0.2.0 (2022-05-18)

New scipts

• New scipt: Optimization_run.py

Caveats

• For now works only with hard-coded paths, so one needs copy-paste supplementary folders and scipts

0.1.3 (2022-05-15)

• Minor changes to package wrappers

0.1.2 (2022-05-15)

• Minor changes to package wrappers

0.1.0 (2022-05-15)

• First release on PyPI.