Change request #1757
cssens should respect previously applied energy cuts (and thresholds)
| Status: | In Progress | Start date: | 04/08/2016 | |
|---|---|---|---|---|
| Priority: | Normal | Due date: | ||
| Assigned To: | - | % Done: | 0% | |
| Category: | - | |||
| Target version: | - | |||
| Duration: |
Description
I want to compute a sensitivity curve for a large observation container. The observation container was preselected according to given energy thresholds.
When running cssens in differential mode, the function set_obs_ebounds simply overwrites the existing energy boundaries of each observation. It should however, respect previous energy cuts and omit observations when their energy range does not contribute in the differential sensitivity bin.
Recurrence
No recurrence.
History
#1
Updated by Mayer Michael almost 8 years ago
- Status changed from New to Pull request
I have implemented the change on branch 1757-cssens-respect-ebounds.
#2
Updated by Knödlseder Jürgen almost 8 years ago
- Tracker changed from Bug to Change request
- Status changed from Pull request to In Progress
Sorry for not looking into this earlier.
I had a look into your code and I was wondering whether it actually does what you want. The point is that _set_obs_ebounds() is called for evergy energy bin [emin,emax], and after the first call to the method the original observation energy boundary information is lost. I think for the code to work properly one would need to store all observation energy ranges somewhere, and use this stored information is every call.
#3
Updated by Mayer Michael almost 8 years ago
Indeed, I haven’t considered that.
I slightly updated the code on the respective branch. The class cssens now contains a new member self._obs_ebounds = [] that stores a list of the individual ebounds objects from the observation container. The list is filled right after the _get_parameters call. At the end of the function _get_sensivitiy(), I restore now all energy boundaries by the following loop:
for i, obs in enumerate(self._obs):
obs.events().ebounds(self._obs_ebounds[i])
I am still wondering if the enumeration is safe enough to ensure the ebounds and observations objects are not mixed up. As an alternative, we could use a Python dictionary that stores energy boundaries by
obs_id or name keys. What do you think?