bcm1futcaprocessor bugs
- if channel is noisy, zero-probability is forced to be =0, giving infinities from the log(), hence causing NaNs in the average luminosity
- beam info is only reset at the processor restart - need to be also done when there is new fill
- orbit normalization assumes that the source is always sending 4NB, which is not true. This causes bad values at new Runs. Number of orbits needs to be obtained from the source.
Edited by Joanna Wanczyk