psKernelGenerate algorithm options/concerns

[robert.desonia@…]

I'm not 100% comfortable with my implementation of psKernelGenerate yet, as I
had to make some assumptions concerning the exact interpretation of the input
vectors that are not trivial.

1) Do I assume the opening of the shutter is at t=0 and the initial x,y is 0,0
or is the first entry of the vectors coorespond to the open-shutter event?

2) Since the time of the close-shutter event is not known otherwise, I'm
assuming that the last time entry is the this time. I, therefore, ignore the
x,y vector values at that index (i.e., a shift coincident with the close-shutter
event doesn't contribute to exposure). Is this acceptable?

Should there be other parameter(s) to specify the exposure time span? This is
definitely dependent on the original form of the data expected from the
observatory software, and since I don't have any interface documentation
detailing this, I worry that I haven't implemented it to interface with the data
as it comes from the mountaintop. For instance, I can easily imagine that the
form of the data from the observatory may actually be in multiple parts: a
shutter-open time, an exposure length, and a set of x,y,t points-in-time
representing shift events in the orthogonal-transfer controller. As
psKernelGenerate is currently specified, a reformat of the data would be required.

-rdd

[Paul Price]

conflat2 by John Tonry

[Paul Price]

Example shifts file from OPTIC

[Paul Price]

Closing subsequent to release of SDRS-08, ADD-07.