Fix the same number of looks values in the GUNW metadata

[xhuang-jpl]

This PR is to fix the same number of looks values for the both wrapped and unwrapped interferogram in the GUNW metadata reported by Ian Joughin. The issue is raised because the GUNW writer class is inherent from both the RIFG and RUNW writer classes where those two variables self.igram_range_looks and self.igram_azimuth_looks are overwritten when the function add_interferogram_to_procinfo_params_group is called by one of the RIFG and RUNW writers.

[gmgunter]

@xhuang-jpl, we briefly reviewed this offline and you gave some context for this change (thank you!):

• The number of looks may differ between the wrapped product (RIFG) and the unwrapped product (RUNW).
• When the number of looks in the phase_unwrap group of the runconfig is set to 1 (in both azimuth and range), then no additional looks are taken in the unwrapping step, so the number of looks in the unwrapped phase is the same as the number of looks in the wrapped interferogram. This is also documented in the runconfig:

isce3/share/nisar/defaults/insar.yaml

Lines 597 to 600 in 32642e4

	# Number of looks in slant range and azimuth directions to generate the
	# wrapped interferogram that will be unwrapped in RUNW.
	# If range_looks = 1 and azimuth_looks = 1 no further multilooking occurs
	# In this case the wrapped interferogram in RIFG will be unwrapped.

Thanks for this clarification. I think the thing I'm still confused about is this:

If the number of looks in the phase_unwrap group of the runconfig is not 1, should it be the total number of looks, or the additional number of looks after multilooking by the number of looks specified in the crossmul group of the runconfig?

[xhuang-jpl]

Thank you @gmgunter, if the number of looks in the phase_unwrap is not 1, then it will be the total number of looks. The unwrapping and crossmul have independent number of looks.

[hfattahi]

Yes I agree that it should be the total number of looks in that situation. The user would only care about the final number of looks and if looks are taken in one step or two steps is just implementation detail.

[hfattahi]

LGTM!

[Tyler-g-hudson]

LGTM

[Tyler-g-hudson]

Suggested change

	unwrap_igram_range_looks = wrap_igram_range_looks
	unwrap_igram_azimuth_looks = wrap_igram_azimuth_looks
	if (unwrap_az_looks != 1) or (unwrap_rg_looks != 1):
	unwrap_igram_range_looks = unwrap_rg_looks
	unwrap_igram_azimuth_looks = unwrap_az_looks
	if (unwrap_az_looks != 1) or (unwrap_rg_looks != 1):
	unwrap_igram_range_looks = unwrap_rg_looks
	unwrap_igram_azimuth_looks = unwrap_az_looks
	else:
	unwrap_igram_range_looks = wrap_igram_range_looks
	unwrap_igram_azimuth_looks = wrap_igram_azimuth_looks

Just a little nitpick - these variables are always going to be assigned either way, this just organizes the assignments in an if/else block for fast reading of the logic