-
…at are not covered by the RoI.
-
Removed all the deprecated code fragments from previous MC generation method.
-
'update_mc_cache()' now initializes an object of type 'GModelSpectralBrokenPlaw' used in MC generation. Removed the entire batch of code which generates from a broken power law and replaced with generation of energies from the GModelSpectralBrokenPlaw::mc() method.
-
# Conflicts: # ChangeLog # NEWS
-
Added gradient formulae. Added helpful notes.
-
Previously neglecting the appropriate handling in the CDF computation of the different indices above and below the break energy. Now appropriately handles the case where (Emax < Ebreak) and (Emin > Ebreak).
-
Both methods for specifying the parameters of a "SmoothBrokenPowerLaw" model in an XML file are now supported. Updated the test to reflect the actual way a spectral model is constructed in GammaLib. The method still has some issues when generating the monte carlo. This is being worked on.
-
Added file "GModelSpectralSmoothBrokenPlaw.i" for python wrappers. Added file "model_point_smoothbplaw.xml" for testing the construction of the class from xml file. Updated "test_GModel.cpp" to include test of the 'GModelSpectralSmoothBrokenPlaw' class. Some work is still needed to ensure the test passes when running 'make check'.
-
Implemented special handling in the MC event generation for when either (or both) the indices are -1. Removed unnecessary vectors used for MC event generation.
-
Implemented Monte Carlo code for generating events from broken power law distribution, then rejecting based on relative values of the power law vs smoothly broken power law. Still need to implement special handling for when one of the indices =-1.
-
…hrough wcs_set (#1851). Formerly the wcs_set() method was called in the derived classes, which led problems with the GWcsGLS class that derives from GWcsSFL. Since GWcsSFL class wcs_set() with parameters that are intended for GWcsGLS, an exception has occured in the initial class. This problem is circumvented by placing the wcs_set call in the base class so that it will only be called once by any of the derived classes.
-
Implemented 'read()' and 'write()' methods.
-
Added methods for numerically computing the flux.
-
Finished implementing gradient formulae for 'pivot', 'breakenergy', and 'beta'. All parameters now have a gradient formula.
-
Added "GModelSpectralSmoothBrokenPlaw.hpp" to "GammaLib.hpp" GModelSpectralSmoothBrokenPlaw.hpp: - Added the 'pivot' parameter and relevant methods. - Added cache parameters for various values to aid in the computation of parameter gradients. GModelSpectralSmoothBrokenPlaw.cpp: - Added 'pivot energy' to model definitions. - Updated documentation text. - Updated '::eval()' method and gradient computations (still need to add gradient formulae for 'pivot', 'breakenergy', and 'beta', next commit)
-
Added methods for setting/getting the 'beta' parameter. Updated the documentation text for several methods in the '.cpp' file Still many methods to be updated before testing can be done.
-
Added ".cpp" and ".hpp" files for new GModelSpectralSmoothBrokenPlaw spectral model class. Method still not fully implemented. Still need to add ".i" files.
-
… of GModelSpectralNodes and GModelSpectralFunc
-
…on older systems, such as CentOS 5)