Testing Cheetah SPS modules with PSS test vectors

The Klotski implementation of SPS has been fixed for the bugs found in earlier PIs (bug tickets are linked to this feature). MR including fixes: https://gitlab.com/ska-telescope/pss/ska-pss-cheetah/-/merge_requests/43 We ran all test vectors in the ProTest SPS feature through the SPS code included in that MR, and plots of the single pulse detections are collected in this document: https://docs.google.com/document/d/1PeSS5OrMvkvSMvijnn_osg-zk71WET5hd_dH6TRF57I/edit?usp=sharing We have also turned on the previously defined single pulse tolerances in ProTest and checked how many pulses are now detected by ProTest. The results are written up in this document: https://docs.google.com/document/d/1SXMo4OZFqfcq3A2N8IyGOvzPCDDg4g7Xs-ihnmg4FHQ/edit?usp=sharing In summary, some of the pulses that ProTest expects are found in SPS outside of the tolerance values: ProTest detects 122 of the 136 pulses we passed through SPS. We believe that the SPS code is now all correct, but some of the tolerance values are too restrictive, in particular the DM tolerance. We will look into further refining the tolerances in future PIs. Ben Shaw also completed some training on pandas covering the basics of series, data frames, and most importantly, conditional filtering (which will allow us to instantly exclude all the candidates which fall outside our tolerances values, on any field). Also covered, various methods for performing statistical analysis on the data, and well as how to handle missing or null vales. We think this will be very useful for ProTest development in the future.