Tracing Quinoa

Main changes:

• Python 2.7 is not supported anymore
• Python 3.6 and 3.7 are supported
• Several changes to try and make installation easier (on clusters, windows, mac....) and less verbose for users
• More explicit names for default saved configurations
• Major bug fix in one of the methods for computing synthetic signal
• Minor bug fixes in interactive figures
• Minor bug fixes in Plasma2D interpolation
• New configuration (ITER) available
• First version of a class handling 2D XRay bragg spectrometers
• First tools for magnetic field line tracing available on WEST
• Better documentation, more ressources
• More informative error messages
• extra tools for computing LOS length, closest point to magnetic axis...
• Better PEP8 compliance

Detailed changes:

Installation / portability:

• Bug fixes for installation on ITER and Gateway clusters #250
• Easier installation on Mac #183 (requirements.txt)
• Removed explicit compiler specification in setup.py for more flexibility
• When sub-packages imas2tofu or mag are missing, the warning is much more concise, but the error traceback is still accessible in a hidden dictionnary for developpers #243
• python setup.py clean now doesn't cythonize
• Dropped python 2 support: #228
  • merged _GG02 and _GG03 into the _GG file
  • no longer being tested in travis
  • no longer packaged in conda
  • updated README file accordingly
  • dropped funcsigs dependency
  • made necessary changes in setup.py
  • benchmarks/calc_signal_benchmark.py: now working with python 3

Bug fixes:

• Major bug fix in LOS_calc_signal() for computing the synthetic signal of a LOS camera using a particular algortihm : method='sum', minimize='hybrid', ani=True, #247 #255
• Major bug fix in LOS_get_sample() when minimize='hybrid' and minimize='memory' the limits were not set correctly
  in some cases the formula for sampling a LOS was wrong (los_get_sample_core_var_res).
• Minor bug fixes in interactive figures when t=None was used (the interactivity was lost due to wrong formatting of the time array)
• Minor bug fixed in Plasma 2D interpolation (interp_t was not being set), imporved error messages #209
• Removed unused variable in _Ves_get_sampleS (_GG), in _core.py #190
• Gave more explicit names to some variables in _core.py to avoid bugs/typos (eg. I to current) #186
• Removed a secondary init function for the class tf.geom.CoilsPF #187
• _checkformat_inputs_dgeom now is a function of Rays class
• Change default separator in to_dict() from '_' to '.', #228
• Rays.calc_kInOut_IsoFlux() is now more flexible (accepts 3d np.arrays as well as lists of np.ndarrays), #188
• repr() is now overloaded in a more robust way so it falls back to printing the class if get_summary() fails to avoid crashing in debug mode #242

Documentation:

• Updated information about support of python version
• Added slides of talk given at PyConFR 2019 conference
• Added a gallery in our documentation with 3 different tutorials:
  • 5 minutes tutorial to show to create a geometry and 1D/2D cameras
  • Guide on how to create your own Geometry from scratch (vacuum vessel, structures, etc.)
  • How to compute the signal received by a camera using a synthetic signal.
    For all of these tutorials, you can see directly the codes and the
    resulting images, and you can get the source code or download it as a
    Jupyter notebook!
• Minor changes to the web doc: updated install instructions to be "cleaner"
  now in rST and not HTML), small changes in the navigation bar.
• Guide on how to contribute to ToFu.

New features:

• First version of magnetic field line tracing (for WEST only so far, to be improved) #213 #235
• First version of 2D XRay Bragg spectrometers for synthetic diagnostics, modelling the diagnotics geometry and providing 2d spectral fitting routines (to be improved) #267
• When computing a signal LOS_calc_signal emissivity function can now return
  a 1D array if t=None #217 #252
• Three functions added to tf.geom.Rays: #211
  • calc_length_in_isoflux(): compute the length inside a set of isoflux surfaces of each LOS
  • calc_min_geom_radius(): compute the minimal geometrical radius (impact parameter) of each LOS
  • calc_min_rho_from_Plasma2D(): compute the minimum normalized radius (or any field with a minimum on the axis) for each LOS
• New ITER configuration available! #220 #227
• More explicit names for default configurations #264

Contributors:

Many thanks to all developpers:

• Didier Vezinet (@Didou09)
• Laura Mendoza (@lasofivec)
• Florian Le Bourdais (@flothesof)
• Jorge Morales (@jmoralesFusion)
• Koyo Munechika (@munechika-koyo)
• Louwrens Van Dellen (@Louwrensth)

What's next (indicative):

• Migrating from nosetests (ongoing for @lasofivec : issues #95 and #232 )
• Easier binary and source installs from pip and conda for all platforms, including unit tests on alla platforms (ongoing for @lasofivec and @flothesof : issue #92 and #259 )
• Solid angles for Volume-Of-Sight and radiative heat loads computation (ongoing for @lasofivec : Issues #71, #72, #73, #74, #75, #76, #77, #78 )
• Tools and classes to handle 2D Bragg X-Ray crystal spectrometer (ongoing for @Didou09 : Issues #202 and #263)
• Generic data classe to incorporate plateau-finding, data analysis and 1d Bayesian fitting routines and classes (ongoing for @Didou09 and @jmoralesFusion and @MohammadKozeiha: issues #208, #260 and #262)
• More general magnetic field line tracing workflow
• Better unit tests coverage
• More complete documentation

List of issues and PR closed:

• Issues: #30, #180, #183, #185, #186, #187, #188, #189, #190, #201, #209, #211, #213, #217, #220, #224, #227, #228, #230, #235, #243, #247, #248, #250, #252, #255, #264
• PR: #173, #175, #179, #181, #182, #184, #191, #192, #193, #194, #195, #196, #197, #199, #206, #207, #210, #212, #222, #223, #225, #226, #229, #231, #233, #234, #236, #237, #238, #240, #242, #244, #245, #246, #249, #251, #253, #254, #256, #257, #258,
  #261, #265, #267