BOZ analysis improvement

• split BOZ part II into:
  • processing and saving DSSC data into xarray small data
  • loading small data, binning multiple runs, plotting results and saving binned dataset
• add to BOZ part I an extension where validity of correction of previous day can be checked on new data (3 broken membranes)
• BAM has 'sa3_pId' while DSSC has 'pulseId' which don't match
• for dark, pedestals is computed as mean, we could compute bounds to find bad pixels as std or percentiles
• improve finding rois with automatic threshold. In many cases the threshold has to be adjusted
• give information on the percentage of saturated shots
• plot an histogram of Io intensity over run
• save corresponding data of all plots