-
Remove ctools::time_reference constant and handle the time reference for MET either by using the time reference from the input observation, or by adding a hidden "mjdref" parameter. The following tools transparently use the time reference of the input observation: * ctselect (apply time reference of each input observation) * ctbin (apply time reference of first input observation) * csobsselect (apply time reference of each input observation) A "mjdref" parameter was added the following tools: * ctobssim * cslightcrv * cspull * csviscube * csobsdef
-
The working log file pointer was not forwarded to ctobssim::set_mc_id_names() and consequently the standard log file pointer was used, which occasionally lead to the exception during the Python unit test. Now the log file pointer is forwarded properly.
-
# Conflicts: # ChangeLog # NEWS
-
All likelihood ctools now derive from the ctlikelihood base class. This avoids code duplication and provides all likelihood tools with a uniform interface.
-
The true energy range was used for event selection instead of the reconstructed energy range. This was a bug introduced in the recent refactoring. Now fixed.