various minor changes, and updated opendap mapper for Arome Arctic forecast data #533
Conversation
…very time a nansat object was opened
… I've changed it to a debug message.
…he export function instead
… issue when fill values are not correctly picked up
…ute names without 'GDAL_'
…'CF-1.5' which is obviously wrong. This is now deleted. Same goes for the history. That should also be handled in the processing software rather then by the gdal netcdf driver.
nansat/exporter.py
Outdated
| @@ -123,25 +129,29 @@ def export(self, filename='', bands=None, rm_metadata=None, add_geolocation=True | |||
| # Rename variable names to get rid of the band numbers | |||
| self.rename_variables(filename) | |||
| # Rename attributes to get rid of "GDAL_" added by gdal | |||
| self.rename_attributes(filename) | |||
| self.correct_attributes(filename, history=self.vrt.dataset.GetMetadata()['history']) | |||
There was a problem hiding this comment.
This makes a lot of unit tests fail. May I suggest the following:
| self.correct_attributes(filename, history=self.vrt.dataset.GetMetadata()['history']) | |
| self.correct_attributes(filename, history=self.vrt.dataset.GetMetadata().get('history')) |
| """ | ||
| _, filename = os.path.split(filename) | ||
| t = datetime.strptime(filename.split('_')[-1], '%Y%m%dT%HZ.nc') | ||
| return datetime.strftime(t, '%Y-%m-%dT%H:%MZ') | ||
| return datetime.strptime(filename.split('_')[-1], '%Y%m%dT%HZ.nc') |
There was a problem hiding this comment.
All other mapper's get_date() method return a formatted string.
I think it would be best to keep consistency across all mappers, would you mind reverting this?
| it as a datetime object. | ||
| """ | ||
| _, filename = os.path.split(filename) | ||
| return datetime.strptime(filename.split('_')[-1], '%Y%m%dT%HZ.nc') |
There was a problem hiding this comment.
Same comment as in mapper_opendap_arome.py: for consistency across mappers, I think it would be best to return a formatted string.
| return metadictlist | ||
|
|
||
|
|
||
| def _get_band_from_subfile(self, fn, netcdf_dim={}, bands=[]): |
There was a problem hiding this comment.
There are still calls to _get_band_from_subfile() in mapper_netcdf_cf_sentinel1.py and in the unit tests (test_mapper_netcdf_cf.py).
Would you mind updating these with your new methods?
| @@ -526,9 +526,9 @@ def _create_extent_dict(extent_str): | |||
| raise ValueError('<extent_dict> must contains exactly 2 parameters ' | |||
| '("-te" or "-lle") and ("-ts" or "-tr")') | |||
| key, extent_dict = Domain._add_to_dict(extent_dict, option) | |||
| if key is 'te' or key is 'lle': | |||
| if key == 'te' or key == 'lle': | |||
There was a problem hiding this comment.
This change highlights a problem in this test:
the arguments to -te are given in the wrong order.
Could you change "-te -180 180 60 90 -ts 500 500" to "-te -180 60 180 90 -ts 500 500"?
Tests failed locally but, as agreed with @akorosov , we test in actions and see how it goes...