TST: specialized OMNI HRO routines need testing

[jklenzing]

• time_shift_to_magnetic_poles
• calculate_imf_steadiness (w/ custom max_wnum)
• Custom warning in calculate_imf_steadiness
• calculate_dayside_reconnection
• calculate_clock_angle

[jklenzing]

Need to evaluate the usage of these routines before writing tests

[jklenzing]

Notes on usage of calculate_imf_steadiness. In pysatNASA 0.0.2, this will autofill a sample rate of 1 if a tag of '' is attached to the instrument. This allows the routine to be run on simulated data attached to a pysat_testing object. The usage of this should be evaluated in the tests.

[jklenzing]

@aburrell, do we want to maintain support for older scipy?

pysatNASA/pysatNASA/instruments/methods/omni.py

Lines 140 to 150 in d5cf79a

	except TypeError:
	pysat.logger.warn(' '.join(['To automatically remove NaNs from the',
	'calculation, please upgrade to scipy 1.4',
	'or newer.']))
	circ_kwargs.pop('nan_policy')
	ca_std = \
	inst['clock_angle'].rolling(min_periods=min_wnum,
	window=steady_window,
	center=True).apply(stats.circstd,
	kwargs=circ_kwargs,
	raw=True)

scipy 1.4 was released in 2019. I think we could add a limit and drop this. Not sure how we would test for this otherwise

[jklenzing]

Also, I don't see how to trigger the following lines without also creating a valueError later on due to non-integer windows. Is this the intended behaviour?

pysatNASA/pysatNASA/instruments/methods/omni.py

Lines 114 to 120 in d5cf79a

	max_wnum = np.floor(steady_window / sample_rate)
	if max_wnum != steady_window / sample_rate:
	steady_window = max_wnum * sample_rate
	pysat.logger.warning(" ".join(("sample rate is not a factor of the",
	"statistical window")))
	pysat.logger.warning(" ".join(("new statistical window is",
	"{:.1f}".format(steady_window))))

[aburrell]

I think dropping the older scipy is fine. Maybe change lines 114-120 to just raise an earlier ValueError, or adjust the input to prevent the ValueError from happening.

[jklenzing]

The valueError comes directly from pandas when the rolling routines are invoked

[jklenzing]

Turns out it was a bug! 🐞