4detector station can and do recieve configs without slave and break

[tomkooij]

Stations with master/slave (4 detectors) can recieve erroneous config messages without slave. (s9 oct 29, s502 oct 11). ref #191

This breaks reconstructions: The api.Station object reads the most recent config, with 2 scintillators. This means api.Station.n_detectors == 2 and len(cluster.station.detectors) == 2

When reconstructing ESD events with 4 detectors, code like dx, dy, dz = self.station.detectors[id].get_coordinates() will obviously throw an IndexError.

As we cannot easily refurse/delete/ignore erroneous configs, we must fix api.Station (get rid of n_detectors???)

[153957]

The issue originates also from the publicdb;
https://github.com/HiSPARC/publicdb/blob/master/publicdb/inforecords/models.py#L252-L271

This is where the number_of_detectors is determined, which is used in the station API response, upon which SAPPHiRE bases its conclusion of the number of detectors.

Maybe that number_of_detectors should not be based on the presence of a slave config, but on data from the station or some other, more consistent/correct, source of information.

[davidfokkema]

Also, aren't there examples of two-detector stations which were upgraded to be four-detector stations? Then the result is dependent on the date. I think the data should be the definitive source of truth.