Not printing empty variable to files

[BaptisteVandecrux]

At KAN_B for example, there is no GPS, no ice temperature, no pressure transducer

We had already removed the variables cc, dlhf, dshf from the tundra sites:
#155

Should we also do the same with t_i and z_pt?
Some tundra sites might have active GPS ?

Same thing for KAN_U that never had a pressure transducer.

[BaptisteVandecrux]

This could be fixed by adding:

        # removing empty variables
        for var in self.L3.data_vars:
            if self.L3[var].isnull().all():
                logger.info(f'removing {var} because empty')
                col_names.remove(var)

right after colname is being defined:

pypromice/src/pypromice/process/aws.py

Lines 149 to 168 in a7997ef

	def writeArr(self, outpath):
	'''Write L3 data to .nc and .csv hourly and daily files
	Parameters
	----------
	outpath : str
	Output directory
	L3 : AWS.L3
	Level-3 data object
	'''
	outdir = os.path.join(outpath, self.L3.attrs['station_id'])
	if not os.path.isdir(outdir):
	os.mkdir(outdir)
	col_names = getColNames(
	self.vars,
	self.L3.attrs['number_of_booms'],
	self.L3.attrs['format'],
	self.L3.attrs['bedrock'],
	)

But we need to decide:

• if there's never been a pressure transducer at a site (like KAN_U) do we want that variable removed?
• if there's been a pressure transducer but it has never been functional (like KPC_Uv3), do we want that variable removed?
• At the moment, for both cases above, the final files contain variables z_pt and z_pt_cor full of NaN. Is that confusing for the data users? If not then I am fine closing this as "won't fix".