Include more uncertainties in A(V) calculation

[mdecleir]

One way to calculate A(V) is from an assumed A(K)/A(V) value.:
Currently, the uncertainty on the A(K)/A(V) value is not taken into account in the uncertainty calculation of A(V).

At the same time, we might want to check if there are more up to date values for A(K)/A(V) in the literature?
Maybe it is possible to use the value from my NIR extinction paper (0.103 from the fit to the average diffuse curve)?

measure_extinction/measure_extinction/extdata.py

Lines 472 to 503 in d13198c

	def calc_AV(self, akav=0.112):
	"""
	Calculate A(V) from the observed extinction curve:
	- fit a powerlaw to the SpeX extinction curve, if available
	- otherwise: extrapolate the K-band extinction
	Parameters
	----------
	akav : float [default = 0.112]
	Value of A(K)/A(V)
	default is from Rieke & Lebofsky (1985)
	van de Hulst No. 15 curve has A(K)/A(V) = 0.0885
	Returns
	-------
	Updates self.columns["AV"]
	"""
	# if SpeX extinction curve is available: compute A(V) by fitting the NIR extintion curve with a powerlaw.
	if "SpeX_SXD" in self.waves.keys() or "SpeX_LXD" in self.waves.keys():
	self.fit_spex_ext()
	# if no SpeX spectrum is available: compute A(V) from E(K-V)
	else:
	dwaves = np.absolute(self.waves["BAND"] - 2.19 * u.micron)
	kindx = dwaves.argmin()
	if dwaves[kindx] > 0.04 * u.micron:
	warnings.warn(
	"No K band measurement available in E(lambda-V)!", stacklevel=2
	)
	else:
	ekv = self.exts["BAND"][kindx]
	self.columns["AV"] = ekv / (akav - 1)