parsotat
commented
1 year ago Solution:
Cancel the analysis of the observation ID.
If running a parallelized analysis, set the number of processes used = 1 and rerun the ba.parallel.batsurvey_analysis command with this new number of processes. Then, the code will get stuck on the troublesome observation ID which the user should take note of.
If running a serial analysis of a number of observation IDs the user should take note of the observation ID that the code is stuck on.
With the troublesome observation ID noted, modify the list of observation IDs that are passed to be analyzed and ensure that the troublesome ID is no longer included. Then continue the analysis.
This solution may need to be implemented iteratively (as was done for the analysis of the MAXI J0637-430 source to exclude all problematic observation IDs.
It has been found that for some observation IDs, the call to batsurvey gets stuck and the batch analysis of many observation IDs may be halted.
This may be related to the combination of the detthresh and detthresh2 parameters and the image cleaning that needs to be done. It is possible for the number of detectors to be enough for the detthresh value to be satisfied but for there to not be enough detectors for the cleaning to be done in all the energy channels. This causes batcelldetect, which does the cleaning, to get stuck.