Hiproc

-------------------

Added
+++++
- fft_clean was added to the console programs list. Still very experimental.

Changed
+++++++
- Updated EDR_Stats to be consistent with upstream EDR_Stats Pipeline version 3.0.0 (2021/09/08).
This handles the new LUT settings.
- Updated HiCal to be consistent with upstream HiCal Pipeline version 4.2.3 (2021/11/17).
- Updated HiStitch to be consistent with upstream HiStitch Pipeline version 2.21.2 (2021/09/09).
This change defaults to balance processing for all CCDs and corrects an indexing bug.

Fixed
+++++
- HiCal now deals with missing fields in the output of histats.
- lisfix now does not try to fit a slope to a column with masked values.
- bitflips now raises more descriptive exceptions when given poor data.

-------------------

Added
+++++
- fft_clean functionality added for more testing, but not integrated with HiCal
or hiproc.

Fixed
+++++
- EDR_Stats was failing when some histats return values were None. Now properly
just saves a "None" for SNR.
- HiCal multiprocessing approach was failing when a Channel was bypassed, essentially breaking the whole run.
- When the HiStitch function was called externally, a missing channel wasn't handled properly. When called from the
HiStitch console script, it was fine.
- hiproc adjusted to take advantage of the above two fixes to HiCal and HiStitch.
- HiccdStitch now tolerant against the new ISIS stats return of the "N/A" string for missing values.

------------------

Added
+++++
- HiJACK now retains pre- and post-HiJACK flat files by default.
- HiJACK now has a --plotflats argument that can be used to plot the new flat files *after*
a run of HiJACK.
- Added new rjplot program to plot the contents of the jitter and smear output files created
by resolve_jitter.

Fixed
+++++
- In certain conditions an array can have zero length, and when it does, one shouldn't try and
index it in a logging statement.

------------------

Added
+++++
- Additional logging of the HiPrecisionInit output during a hiproc run, which was previously
obscured.

Fixed
+++++
- In some legitimate cases, some elements returned from ISIS histat would be missing (e.g. in the
case of no valid pixels in the CAL_BUFFER area, there is no returned value of "Average"), but
were not handled properly (Issue 3). More robust handling was added.

------------------

Changed
+++++++
- bitflips.py: Removed the span check when considering end values, which may affect how
the algorithm deals with "mostly good" DN histograms, such that it should now correctly
handle mesa-shaped DN histograms and not chop them in the middle.

------------------

Fixed
+++++
- The new-in-0.7.0 multiprocessing requires that the multidict package be installed
(for pvl to use), and it wasn't included in the requirements. If you have 0.7.0,
a simple "conda install multidict" should get you working.