Try using the `hope` framework to optimize parts of the code.
hope
is a python package for runtime compilation of Python code into C/C++ code. We can potentially speed up some methods, especially stuff that is associated with Likelihood evaluations and get_detector_response
.
-
Identify parts of the code that could benefit from those adaptations -
Refactor functions so that it can be compiled by hope
(see http://hope.readthedocs.io/en/latest/lang.html) -
Do profiling runs to see how much we gain by implementing hope.