Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>

* Class field added to some old element creation functions

* wiggler.m old file compatibility with atwiggler

* update old element creation as in quadrupole.m

* energy variable in wiggler.m different options

* Unify the settings of differentiation steps, improved atlinopt

* Improved Matlab help

* Improved Matlab help

* Added system and numpy info to pytest report

* Fixed system info on Windows

* Use platform.uname instead of os.uname()

* corrected scaling

* corrected scaling

* Last 2 scalings corrected

* scalings in findorbit6

* Use Github Actions for continuous integration.

This provides builds on Linux, Windows and Mac. It does not provide the
capability of building and uploading wheels to PyPI as yet.

* Temporary physics tests

* Temporary physics tests

* Removed "missing" variables for Octave compatibility

* Added keyword arguments for XYStep and DPStep where missing

* Changed default steps to XYStep=3e-8, DPStep=3e-6

* Added comments on the optimization of XYStep and DPStep

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>
Co-authored-by: Will Rogers <will.rogers@diamond.ac.uk>

* Removed elementpassdemo, identical to trackingdemo

* Fixed atenergy, FODO

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>

Try creating Github Action.

This provides builds on Linux, Windows and Mac. It does not provide the
capability of building and uploading wheels to PyPI as yet.

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>

* Wiggler field, wiggler integrals

* Correct global variable name (#187)

* corrected Bwig

* Bug fixes

* Bug fixes

* Bug fixes

* Updated atsummary and removed debug info

* Analytical I3

* amat, atradon

* Fixed missing 'beta' in sync tune and bunch length

* Fixed missing 'beta' in sync tune and bunch length

* Removed temp files

* Fix import of wigglers from matlab

* Check overvoltage < 1

* Wiggler radiation in python

* Remove useless import, bug fix

* Improved docstring

* vector ordering in amat

* fix coupled tunes

* Handle wigglers in atradon and atradoff

* Ignore wigglers using DriftPass

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>

modified gen_detuning_elem to make it coherent with the matlab implem…

Twissline

* Unify the computation of radiation integrals in DipoleRadiation.py

* Pyat metadata.

* Bug correction in atcavityon and atcavityoff (#167)

* First version of amplitude detuning. Definitely WIP and needs major revision but does work

* Added M66 element generation script

* Version Minus1 of fast_ring.

* Missed the actual fast_ring file that I made.

* alternative detuning function

* added alternative detuning

* Some local changes

* New amplitude detuning module

* merge problems

* Readded gen nonlin elem after the merge. Should be good now

* reorganized imports and use jmat from amat

* diffusion matrix implemented in radiation.py

* implemented diffusion matrix

* simplified coding

* element_generation

* pep8 standard

* removed feature from fastring

* remove fastring

* added test for quantdiff

* repeated import removed

* remove lattice attributes in nonlinear.py

* Make lmat fortran array

* bug correction in atcavityon and atcavityoff

* Update atcavityoff.m

* Update atcavityoff.m

* Update atcavityon.m

* Update atcavityoff.m

Rrmoved unused second input argument

* Update atcavityon.m

header harmonized

* Update athelp.m

Correct error with the filename to save doc of at.
Use now a variable to define the name of the file to save the html doc in order to be consistent througout the code.
athelp.mat

* Correct atradon/off

* atcavityon/off calls atradon/off

* avoid duplicated code

* small bug correction in atradoff

* radiation_on / radiation_off modified

* Use correct filename for athelp in .gitignore.

* Update pyat/at/physics/radiation.py

Co-authored-by: Laurent Farvacque <laurent.farvacque@esrf.fr>

* double quotes docstrings

* fix set_value refpts not working for increment=False, val=scalar and nrefpts>1

* atradon: use keywords for sextupole; octupole, wigglers

* bug correction in atcavityon and atcavityoff

* Update atcavityoff.m

* Update atcavityoff.m

* Update atcavityon.m

* Update atcavityoff.m

Rrmoved unused second input argument

* Update atcavityon.m

header harmonized

* Correct atradon/off

* atcavityon/off calls atradon/off

* avoid duplicated code

* small bug correction in atradoff

* radiation_on / radiation_off modified

* atradon: use keywords for sextupole; octupole, wigglers

Co-authored-by: Lee Carver <carver@rnice9-0308.esrf.fr>
Co-authored-by: Simon White <swhite@esrf.fr>
Co-authored-by: Lee Carver <carver@raki2.esrf.fr>
Co-authored-by: Laurent S. Nadolski <nadolski@synchrotron-soleil.fr>
Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>
Co-authored-by: Nicola Carmignani <carmignani@maccarmignani.esrf.fr>
Co-authored-by: Will Rogers <will.rogers@diamond.ac.uk>
Co-authored-by: Laurent Farvacque <laurent.farvacque@esrf.fr>

* Unify the computation of radiation integrals in DipoleRadiation.py

* Restore the Matlab tests

* Removed obsolete comments in the help of ringpara.m

* Fix Appveyor builds on Windows.

Setuptools v50.0.0 introduces a change that broke our build. It's
possible to undo that change with an environment variable but we
really need to understand it and fix it correctly.

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>
Co-authored-by: Will Rogers <wahrogers@gmail.com>
Co-authored-by: Nicola Carmignani <nicola.carmignani@esrf.fr>
Co-authored-by: Lee Carver <carver@rnice9-0308.esrf.fr>
Co-authored-by: Simon White <swhite@esrf.fr>
Co-authored-by: Lee Carver <carver@raki2.esrf.fr>
Co-authored-by: Laurent S. Nadolski <nadolski@synchrotron-soleil.fr>
Co-authored-by: Nicola Carmignani <carmignani@maccarmignani.esrf.fr>
Co-authored-by: Will Rogers <will.rogers@diamond.ac.uk>

* Compute A,B,C,g in linopt only for coupled case

* Put back g=1 and C=0 for intial ld

* intermediate commit get_chrom not working

* initial proposal for twissline

* small correction

* tune not properly set for refpts=[]

* fixed strong output

* removed some unnecessary tests, orbit not taken from twiin

* fixed pep8 and help

* changed twiin by twiss_in

* re-introduced get_twiss() for testing and remove unnecessary tests for linopt with uncoupled lattice

* remove get_twiss

* added tests to compare line with periodic solution

* updated matlab_tests

* Remove obsolete comment.

* added quantdiff matlab comparison

Co-authored-by: Will Rogers <will.rogers@diamond.ac.uk>

Non linear

Matlab tests

Updates for the wheel building code.

* Simplify the argument list of atwiggler

* Documentation on GWigSimplecticPass

* Adapt wiggler tests

* Modified keyword names

* Corrected help string for the Wiggler

Co-authored-by: Laurent Farvacque <laurent.farvacque@gmail.com>

Setuptools v50.0.0 introduces a change that broke our build. It's
possible to undo that change with an environment variable but we
really need to understand it and fix it correctly.

Atbeam