longitudinal-tomography 3.4.2.dev0

A python library for longitudinal particle beam tomography

Copyright 2020 CERN. This software is distributed under the terms of the GNU General Public Licence version 3 (GPL Version 3), copied verbatim in the file LICENCE.txt. In applying this licence, CERN does not waive the privileges and immunities granted to it by virtue of its status as an Intergovernmental Organization or submit itself to any jurisdiction.

INSTALLATION

The computationally intensive or time-critical parts of the library is written in C++ and python bindings are provided using pybind11. The installation and usage of the library is the same for all operating systems, but different dependencies are needed for different operating systems.

Prerequisites

Linux

You need a C++ compiler like g++ installed. This is not required if installing a prebuilt package from acc-py or pypi.

Windows

On Windows computers MSVC >= 14.0 with the Windows 10 SDK is required.

In MinGW and WSL environments the standard g++ compiler works out of the box.

MacOS

No offical tests have been done on MacOS, but presumably g++, clang/llvm should work.

Install

The Longitudinal Tomography package is available in prebuilt wheels for Python 3.6-3.9 on CERN Acc-Py and pypy.org as longitudinal-tomography. The package can thus easily be installed on a Linux machine using

pip install longitudinal-tomography

The package can be installed on a MacOS or Windows machine in the same manner, but the C++ extension will be built on install.

Other ways to install

Clone the repository and run

pip install .

The C++ extension will be built on install.

For development environments where it’s preferable to compile the C++ extension inplace, it’s possible to run the command

pip install -e .

which will compile the C++ extension using the available compiler (decided by setuptools).

Documentation

This development is based on the well tested and widely used FORTRAN95 code, documented and available here: http://tomograp.web.cern.ch/tomograp/ Details on the algorithms in both codes, and the differences between them, can be found here: https://cdsweb.cern.ch/record/2750116?ln=ka

Parallelization using OpenMP

The C++ extension is accelerated by OpenMP parallel for loops. It is possible to limit the number of launched threads by setting it in the extension, by

from longitudinal_tomography.cpp_routines import libtomo
libtomo.set_num_threads([num_threads])

which will set the maximum number of used threads to [num_threads].