def main(cfg):
"""Run the diagnostic."""
sns.set(**cfg.get('seaborn_settings', {}))
ecs_file = io.get_ancestor_file(cfg, 'ecs.nc')
tcr_file = io.get_ancestor_file(cfg, 'tcr.nc')
ecs_cube = iris.load_cube(ecs_file)
tcr_cube = iris.load_cube(tcr_file)
# Project
if (ecs_cube.attributes.get('project', 'a') != tcr_cube.attributes.get(
'project', 'b')):
raise ValueError(
"ECS and TCR input files have either no 'project' attribute or "
"differ in it")
project = ecs_cube.attributes['project']
# Remove missing data and use equal coordinate
[ecs_cube, tcr_cube
] = iris_helpers.intersect_dataset_coordinates([ecs_cube, tcr_cube])
# Create plot
plot_path = plot_data(cfg, ecs_cube, tcr_cube)
# Write netcdf file
netcdf_path = write_data(cfg, ecs_cube, tcr_cube)
# Provenance
ancestor_files = [ecs_file, tcr_file]
provenance_record = get_provenance_record(project, ancestor_files)
provenance_record.update({
'plot_file': plot_path,
'plot_types': ['scatter'],
})
with ProvenanceLogger(cfg) as provenance_logger:
provenance_logger.log(netcdf_path, provenance_record)
def test_get_ancestor_file(mock_get_all_ancestors, files, output):
"""Test retrieving of single ancestor file."""
mock_get_all_ancestors.return_value = files
if isinstance(output, type):
with pytest.raises(output):
io.get_ancestor_file(CFG, pattern='*')
return
returned_file = io.get_ancestor_file(CFG, pattern='*')
assert returned_file == output
def test_get_ancestor_file(mock_logger, mock_get_all_ancestors, files, output,
logger):
"""Test retrieving of single ancestor file."""
mock_get_all_ancestors.return_value = files
returned_file = io.get_ancestor_file(CFG, pattern='*')
assert returned_file == output
if logger:
mock_logger.warning.assert_called()
else:
mock_logger.warning.assert_not_called()
def get_external_cubes(cfg):
"""Get external cubes for psi, ECS and lambda."""
cubes = iris.cube.CubeList()
for filename in ('psi.nc', 'ecs.nc', 'lambda.nc'):
filepath = io.get_ancestor_file(cfg, filename)
cube = iris.load_cube(filepath)
cube = cube.extract(
ih.iris_project_constraint(['OBS'], cfg, negate=True))
cubes.append(cube)
cubes = ih.intersect_dataset_coordinates(cubes)
return (cubes[0], cubes[1], cubes[2])
def main(cfg):
"""Run the diagnostic."""
sns.set(**cfg.get('seaborn_settings', {}))
input_data = cfg['input_data'].values()
project = list(group_metadata(input_data, 'project').keys())
project = [p for p in project if 'obs' not in p.lower()]
if len(project) == 1:
project = project[0]
# Check if tas is available
if not variables_available(cfg, ['tas']):
raise ValueError("This diagnostic needs 'tas' variable")
# Get ECS data
ecs_filepath = io.get_ancestor_file(cfg, 'ecs.nc')
ecs_cube = iris.load_cube(ecs_filepath)
# Create iris cubes for each dataset
hist_cubes = {}
pi_cubes = {}
for data in input_data:
name = data['dataset']
logger.info("Processing %s", name)
cube = iris.load_cube(data['filename'])
# Preprocess cubes
cube.convert_units(cfg.get('tas_units', 'celsius'))
cube = cube.collapsed(['time'], iris.analysis.MEAN)
# Save cubes
if data.get('exp') == 'historical':
hist_cubes[name] = cube
elif data.get('exp') == 'piControl':
pi_cubes[name] = cube
else:
pass
# Plot data
plot_path = plot_data(cfg, hist_cubes, pi_cubes, ecs_cube)
# Write netcdf file
netcdf_path = write_data(cfg, hist_cubes, pi_cubes, ecs_cube)
# Provenance
ancestor_files = [d['filename'] for d in input_data]
ancestor_files.append(ecs_filepath)
provenance_record = get_provenance_record(project, ancestor_files)
if plot_path is not None:
provenance_record.update({
'plot_file': plot_path,
'plot_types': ['scatter'],
})
with ProvenanceLogger(cfg) as provenance_logger:
provenance_logger.log(netcdf_path, provenance_record)
def _get_style(dataset_name, cfg):
"""Get style for individual data points."""
style = plot.get_dataset_style(dataset_name, cfg.get('dataset_style'))
if not cfg.get('marker_file'):
return style
marker_file = os.path.expanduser(cfg['marker_file'])
if not os.path.isabs(marker_file):
marker_file = io.get_ancestor_file(cfg, marker_file)
data_frame = pd.read_csv(marker_file)
marker_column = cfg['marker_column']
for column in ('dataset', marker_column):
if column not in data_frame.columns:
raise ValueError(
f"Marker file '{marker_file}' does not contain necessary "
f"column '{column}'")
marker = data_frame[marker_column][data_frame['dataset'] == dataset_name]
if len(marker) != 1:
raise ValueError(
f"Expected exactly one entry for marker of '{dataset_name}' in "
f"file '{marker_file}', got {len(marker):d}")
style['mark'] = marker.values[0]
return style