| Version 9 (modified by , 6 years ago) ( diff ) |
|---|
Table of Contents
This is an overview of the most important tables and some important remarks. Links are given to the comments and type of each column.
RunType
Source
RunInfo
The RunInfo table is the most important one. It stored run-wise information. In a first step, a lot of data is directly extracted from the header of the raw run-files. Later, values are calculated from other sources, mainly slow-control file (auxiliary data: aux) and filled into the table as well.
Calculating the on-time
The best estimate for the on-time of the data acquisition is the difference between the first and last event (fRunStop-fRunStart). It should be precise with roughly the data rate, so better than 1s. The on-time calculated by the FTM is only precise with the FTM rate which is a few seconds, but the FTM provides accurate information on the relative on-time w.r.t. to the total time. This information can be used to calculate the on-time in seconds of a run as follows:
TIME_TO_SEC(TIMEDIFF(fRunStop,fRunStart))*fEffectiveOn
For a 1min run, the accuracy should be in the order of 1s/60s (<2%) and for a 5min run 1s/300s (<0.4%). These are upper limits, as in reality, the rate of events is usually higher than 1Hz and the timing of the Ethernet connection should be at least in the order of the average time-difference between two events as well (although this is not guaranteed).
Selection by zenith angle
Ideally, if you select events from a certain zenith angle range (e.g. 10°-50°), you don't want events outside of that range. So, ideally the lower bound is cut on fZenithAngleMin and the upper bound on fZenithAngleMax like
fZenithAngleMin>10 AND fZenithAngleMax<50
Threshold setting
To select data by threshold, the ideal values is usually fThresholdMinSet. It reflects the value obtained as minimum threshold during a run from the currents at its beginning. While some trigger patches could have a significantly higher threshold during a run (which affects the mean value over all patches) the MinSet value reflects usually the majority of the patch thresholds and should be similar (or in most cases identical) to the average (over all events) median (over the camera).
Events
--- not yet available ---
Position
--- not yet available ---
AnalysisResultsRunISDC
factdata/AnalysisResultsRunISDC
The number of background events (fNumBgEvts) is already divided by the number of wobble regions (5).
To calculate the signifciance, you can simply do LiMa(fNumSigEvts, fNumBgEvts). To calculate the statistical error, use ExcErr(fNumSigEvts, fNumBgEvts)`. Both function assume that the number of background events given is the average number and that the scale factor is 5.
