Event Class Handling

Pass 5 and Prior Implementations (ST v9r4p2 and earlier)

• Prior to DC2, there was a separate event class designation for each section of the LAT where the event converted, e.g., DC1::FRONT, DC1::BACK were two distinct classes for otherwise identical event selections. These were given values of 0 and 1, respectively, in the EVENT_CLASS column of FT1. The CONVERSION_TYPE column was set to have the same values, indicating a front (thin) or back (thick) section conversion.
• For DC2, there were two distinct sets of cuts that divided the events into two classes, A and B. This resulted in 4 possible values for the EVENT_CLASS column, and two values for CONVERSION_TYPE:

  event class name	EVENT_CLASS	CONVERSION_TYPE
  DC2::FrontA	0	0
  DC2::BackA	1	1
  DC2::FrontB	2	0
  DC2::BackB	3	1

  Besides still containing redundant information, this required special code in gtselect and the Likelihood tools to handle.
• For Pass 5, Bill defined a new classification tree variable, CTBClassLevel, and proposed three different cuts corresponding to what he envisaged as the corresponding kind of science analysis:

  CTBClassLevel cut	science analysis designation
  > 0	"transient"
  > 1	"source"
  > 2	"diffuse"

  Even though the CTBClassLevel value did partition the event data just as the DC2 class A and class B designations did, the mechanism for handling classes that was used for DC2 was unwieldy, and so it was not modified to handle these new classes. A temporary ad hoc procedure was adopted in which a CTBCLASSLEVEL column was added to the FT1 files (outside of the FFD definition), and fcopy or some other tool besides gtselect was used to make the CTBCLASSLEVEL cuts. The EVENT_CLASS and CONVERSION_TYPE columnss simply indicated front (0) vs back (1), and three separate sets of IRFs were generated for each of the cuts. The principal disadvantage of this scheme is that gtdiffrsp needs to be run separately for each CTBCLASSLEVEL cut, and there was no DSS keyword handling implemented to account for which cut was applied for any given dataset.

Proposed Scheme (25 Mar 2008)

• We will use the EVENT_CLASS column to designate the single event class to which are particular event belongs. In the case of Pass 5, the CTBCLASSLEVEL values can be used. Separate front vs back conversion type information will not be encoded in the EVENT_CLASS value, but it will instead just reside in the CONVERSION_TYPE column, thereby factoring out the redundant information that had been contained in previous implementations. This scheme will have several consequences:
  • Separate IRFs for each class will need to be generated, but instead of generating the IRFs using events that have a minimum CTBClassLevel, as we currently do, the IRFs will need to be generated for significantly smaller numbers of events for which CTBClassLevel is a specific value, e.g., CTBClassLevel==1, CTBClassLevel==2, CTBClassLevel==3. In this case, only the IRFs for CTBClassLevel==3 would be the same as an original Pass 5 set (diffuse class). There would be substantially fewer events for CTBClassLevel==0 or 1. Here is a plot of the effective area for the front section for the three CTBClassLevel values using the AllGamma_v13r9p3_Lyon data:
    

    CTBClassLevel(s)	curve	number of events
    1	red	0.27 M
    2	green	0.24 M
    3	blue	4.30 M
    > 0	black	4.81 M

  • Since each event has a unique set of IRFs,