cpymad 0.14.0

Cython binding to MAD-X

CPyMAD is a Cython binding to MAD-X.

MAD-X is a software package to simulate particle accelerators and is used at CERN and all around the world. It has its own proprietary scripting language and is usually launched from the command line.

This version of CPyMAD is tested with MAD-X 5.02.08. Other MAD-X versions (and immediate revisions) might work too, but are more likely to cause problems.

The installable wheel archives that are provided for some versions of python to simplify the installation on windows contain a precompiled version of CPyMAD that is statically linked against MAD-X 5.02.08.

Disclaimer

This is a heavily modified fork of the cern-cpymad package. The fork is not authored or maintained by CERN members.

CPyMAD links against an unofficial build of MAD-X that is not supported by CERN, i.e. this binary may have problems that the official binary does not have and vice versa. This means:

• Only report issues to CERN that can be reproduced with their official command line client.

• Only report issues here if they cannot be reproduced with their official command line client.

See Reporting issues.

Project pages

• Installation

• Source code

• Documentation

• Issue tracker

• Releases

Usage

The Madx class provides a basic binding to the MAD-X interpreter:

from cpymad.madx import Madx

# Start a MAD-X interpretor. All MAD-X commands issued via cpymad will
# be logged to `command_log`:
madx = Madx(command_log="log.madx")

# Show the version of MAD-X that is actually loaded:
print(madx.version)

# Execute one of your predefined MAD-X files:
madx.call('/path/to/some/input_file.madx')

# Only a handful of MAD-X methods are exposed as methods. For others,
# you can use the `command` attribute. For example, to set a beam:
madx.command.beam(sequence='myseq1', particle='PROTON')

# Calculate TWISS parameters:
twiss = madx.twiss(sequence='LEBT',
                   betx=0.1, bety=0.1,
                   alfx=0.1, alfy=0.1)

# Your own analysis below:
from matplotlib import pyplot as plt
plt.plot(twiss['s'], twiss['betx'])
plt.show()

There are alternative syntaxes for the extreme cases where you need more fine grained control over the command string composition or where command fails to do the right thing:

# can't use `global` as attribute, since it's a python keyword:
madx.command.global_(sequence='cassps', Q1=26.58)

# can't use `: ` as attribute:
madx.command('QP: QUADRUPOLE', AT=2, L=1)

# issue a plain text command, don't forget the semicolon!
madx.input('FOO, BAR=[baz], QUX=<NORF>;')

See http://hibtc.github.io/cpymad for further documentation.

Known issues

On windows with python3.3, there is currently no satisfying way to close file handles in the MAD-X process or prevent them from being inherited by default. You have to make sure on your own that you close all file handles before creating a new cpymad.madx.Madx instance!

Hacking

Try to be consistent with the PEP8 guidelines. Add unit tests for all non-trivial functionality. Dependency injection is a great pattern to keep modules testable.

Commits should be reversible, independent units if possible. Use descriptive titles and also add an explaining commit message unless the modification is trivial. See also: A Note About Git Commit Messages.

Reporting issues

If you have a problem with a sequence file, first try to check if that problem remains when using the MAD-X command line client distributed by CERN, then:

• Report the issue to CERN only if it be reproduced with their official command line client.

• Report the issue here only if it cannot be reproduced with their official command line client.

For issues regarding the cpymad code itself or usage information, I’m happy to answer. Just keep in mind to be precise, specific, concise and provide all the necessary information.

See also:

• Short, Self Contained, Correct (Compilable), Example

• How to Report Bugs Effectively

• How To Ask Questions The Smart Way.

Tests

Currently, tests run on:

• The Travis CI service is mainly used to check that the unit tests for pymad itself execute on several python versions. Python{2.7,3.3} are supported. The tests are executed on any update of an upstream branch. The Travis builds use a unofficial precompiled libmadx-dev package to avoid having to rebuild the entire MAD-X library on each invocation.

  

License

CPyMAD must be used in compliance with the licenses as described in the following sections:

License for CPyMAD source

applies to the python source of the CPyMAD package:

Copyright (c) 2013-2015, HIT - Heidelberg Ion-Therapy Center

To the extent possible under law, the person who associated CC0 with
CPyMAD has waived all copyright and related or neighboring rights to
CPyMAD.

You should have received a copy of the CC0 legalcode along with this
work. If not, see <http://creativecommons.org/publicdomain/zero/1.0/>.

Note that this package is a heavily modified fork of the original cern-cpymad package developed at CERN:

Copyright (c) 2011, CERN. All rights reserved.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use cern-cpymad except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

For details, see the version history or ask the package maintainer.

License for MAD-X

applies to binary distributions that do not require a separate installation of MAD-X, i.e. binaries that are statically linked to MAD-X:

CERN

EUROPEAN ORGANISATION FOR NUCLEAR RESEARCH


Program name:                 MAD --- Methodical Accelerator Design

CERN program library entry:   T5001

Authors or contacts:          [e-mail]
                              BE-ABP Group
                              CERN
                              CH-1211 GENEVA 23
                              SWITZERLAND


Copyright CERN, Geneva 1990 - Copyright and any other appropriate legal
protection of this computer program and associated documentation reserved
in all countries of the world. Organisations collaborating with CERN may
receive this program and documentation freely and without charge. CERN
undertakes no obligation for the maintenance of this program, nor
responsibility for its correctness, and accepts no liability whatsoever
resulting from its use. Program and documentation are provided solely for
the use of the organisation to which they are distributed. This program
may not be copied or otherwise distributed without permission. This
message must be retained on this and any other authorised copies. The
material cannot be sold. CERN should be given credit in all references.

Changelog

(dates are in the form dd.mm.yyyy)

0.14.0

Date: 04.03.2016

• Add function to obtain transfer map

• Fix bug with expanded_elements listing too many elements (leading elements were re-listed at the end)

0.13.0

Date: 24.01.2016

• Update to MAD-X 5.02.08:

  • official support

  • automatic tests

  • prebuilt binaries on PyPI (for windows)

0.12.2

Date: 30.10.2015

• Strip trailing underscore from MadxCommands attribute names. This allows the syntax to be used even for python keywords, e.g. m.command.global_()

• Change the behaviour of Madx.globals:

  • when setting string values, set the variable as deferred expression

  • when getting deferred variables, return instances of type Expression

  • when iterating, only show non-constant globals

0.12.1

Date: 13.10.2015

• fix crash due to incorrect parameter name for logging.basicConfig

• fix crash due to missing subprocess.MAXFD on python3.5

• fix coverage report submitted to coveralls.io

0.12.0

Date: 05.10.2015

• expose directory of global MAD-X variables as Madx.globals

• expose directory of global MAD-X elements as Madx.elements

• fix a bug with Elements.__contains__ reporting yes incorrectly

• list only those column of a table that are marked for output

• add function to get row names of a table

0.11.0

Date: 03.07.2015

• Remove models + resource handling from cpymad. If you need these, check them out from the previous version and maintain them in your own code base. This way you are much more flexible to adapt models to your needs.

0.10.8

Date: 02.07.2015

• Public element names are now of the form “foo[3]” or simply “foo”. The syntax “foo:d” can not be used anymore (this form is used by MAD-X only internally and converted at the lowest wrapper level).

• Fix exception when not specifying sequence name on Madx methods

0.10.7

Date: 21.06.2015

• allow redirection of MAD-X standard I/O via Madx constructor

0.10.6

Date: 29.05.2015

• add csv() method for ResourceProvider

• use C loader from yaml for performance if available

• convert madx.metadata.get_copyright_notice

• add accessors to real sequence + elements for model.Sequence

0.10.5

Date: 25.05.2015

• add MAD-X specific metadata in cpymad.madx.metadata

• speedup Travis testing (using caches and docker containers)

0.10.4

Date: 22.04.2015

• prevent MAD-X process from exiting on Ctrl-C (this was an especially nasty feature when using the interactive python interpretor)

• upgrade to MAD-X 5.02.05 (development release from 10.04.2015)

• fix leakage of open file handles into remote process on py2/windows

0.10.3

Date: 29.03.2015

• make sequence.elements.index more convenient: can now handle names with or without the ‘:d’ suffix as well as the special names ‘#s’ and ‘#e’

0.10.2

Date: 05.03.2015

• add some utility functions to work with MAD-X element names and identifiers

• add a setter method for values to Madx

• improve install instructions. In particular, recommend WinPython as build environment

• fix the MinGW build error due to broken sysconfig inline

• run setup only if invoked as main script

0.10.1

Date: 09.01.2015

• convert IOError to RemoteProcessCrashed, which can occur on transmission if the remote process is already down

• convert ValueError to RemoteProcessClosed, which can occur on transmission if the remote process was already closed

0.10.0 Fork

Date: 09.01.2015

This is the first independent version released for the HIT cpymad fork. The changes on the public API are so heavy, that this is basically a new library.

• rename package from cern.cpymad to cpymad

• remove LHC models from repository

• redesign API to make more use of OOP (no stable API yet!)

• removed some obsolete / unused modules

0.9

Date: 17.11.2014

• don’t link against numpy anymore (this makes distribution of prebuilt binaries on windows actually useful)

• add MAD-X license notice (required to distribute binaries)

• setup.py doesn’t require setuptools to be pre-installed anymore (if internet is available)

• some doc-fixes

• convert cpymad._couch to a simple module (was a single file package)

• use logging through-out the project

• alow logger to be specified as model/madx constructor argument

• multi-column access, e.g.: table.columns['betx','bety']

• move tests one folder level up

0.8

Date: 30.06.2014

• isolate cpymad: remove jpymad backend, remove pymad base

• bootstrap the dependency on numpy

• remove custom MAD-X path discovery during setup. You should use –madxdir if the library is not installed in a system location.

• add function libmadx.is_expanded

• add function libmadx.chdir

• handle MAD-X table columns with integer arrays

• make madx.command more powerful (allows **kwargs and attribute access)

• use inherited pipes for IPC with remote MAD-X processes (allows to forward stdin/stdout separately)

• close connection to remote process on finalization of LibMadxClient

• remove MAD-X command checks, recursive_history and filename completion

• fix name clash

• fix some bugs

• rename convenience constructors to cern.cpymad.load_model and cern.cpymad.start_madx due to name clash with module names

0.7

Date: 16.04.2014

• close handles in remote process properly on all supported python versions

• rewrite libmadx.get_table functionality

• madx functions that return tables now return proxy objects instead. For backward compatibility these can be iterated to allow unpacking into a tuple

• the returned table columns is now a proxy object as well and not TfsTable

• remove retdict parameter

• move some cpymad specific functionality into the cpymad package

• add libmadx/madx functions to access list of elements in a sequence

0.6

Date: 17.03.2014

• raise exception and don’t hang up anymore, if libmadx process crashes

• on python>=3.4, close handles in remote process properly

• let every ‘Madx’ instance have an independent copy of the madx library. this makes the madx module much more useful. previously, this was only true for instances of ‘cpymad.model’.

• restrict to only one cython module that links to libmadx. (allows static linking which is advantageous on windows!)

• use YAML model files instead of JSON

• make ‘madx’ a submodule of ‘cpymad’

• fix test exit status

0.5

Date: 21.01.2014

• migrate to setuptools from distutils

• python3 support

• add continuous integration with Travis

• proper setup.py and MANIFEST.in to be used with PyPI

• rename package to ‘cern-pymad’

• allow to build from PyPI without having cython