GWOLLUM's friend: G2G

G2G stands for Glitch-to-Glitch. This algorithm was designed to isolate and characterize correlations between two trigger samples.

Index:

  1. The principle.

The principle.

The G2G algorithm must be given two trigger samples: the source trigger sample and the target trigger sample. The triggers are first clustered in time. Then a time coincidence operation is performed between the two cluster samples: a coinc event is defined as the time overlap between a source cluster and a target cluster. Three fundamental variables are measured:

  1. The source cluster time: sum of all source cluster durations
  2. The target cluster time: sum of all target cluster durations
  3. The coinc overlap time: sum of all overlaps between the source and target clusters

From these variables, one derives:

  1. The use-time percentage: ratio between the coinc overlap time and the source cluster time
  2. The time efficiency: ratio between the coinc overlap time and the target cluster time

The source and target channels are strongly coupled if both of these variables take high values (close to 100%). However, having high values of the use-time percentage is sufficient to claim a glitch-to-glitch coupling: source → target. In that case, the source channel is said to be a reliable "witness" channel. The time efficiency variable is then used to measures how much the target channels is affected by the noise to which the source channel is sensitive.

The picture below represents how the G2G variables are computed.


Florent Robinet
Contact: robinet@lal.in2p3.fr