Enable use of first group

[stscijgbot-jp]

Issue JP-3891 was created on JIRA by Timothy Brandt:

The zero frame is sometimes separately downlinked with NIRCam data having more than one read per group.  The pipeline does not use this frame for either jump detection or ramp fitting, since it is not evenly spaced in time with the other groups, and it is not independent of the first group.  It can be used with the likelihood approach, however, and can give much better results with four-group data where jump detection is currently treated as an edge case with assumptions about the jump sign.  This requires removing the zero frame from the first group, incrementing the group numbers, inserting the zero frame as the new first group, and modifying the array of read times used by the likelihood fitting algorithm.  

[stscijgbot-jp]

Comment by David Law on JIRA:

Tagging Bryan Hilbert and Alicia Canipe to please weigh in as part of b12.0 planning.