def populate(self):
"""
Compute value count using the objects value count parameters.
"""
# Update status
self.status = self.COMPUTING
self.save()
try:
# Compute aggregate result
agg = Aggregator(
layer_dict=self.layer_names,
formula=self.formula,
zoom=self.zoom,
geom=self.aggregationarea.geom,
acres=self.units.lower() == 'acres',
grouping=self.grouping,
)
aggregation_result = agg.value_count()
self.stats_min, self.stats_max, self.stats_avg, self.stats_std = agg.statistics()
# Track cumulative data to be able to generalize stats over
# multiple aggregation areas.
self.stats_cumsum_t0 = agg._stats_t0
self.stats_cumsum_t1 = agg._stats_t1
self.stats_cumsum_t2 = agg._stats_t2
# Convert values to string for storage in hstore
self.value = {k: str(v) for k, v in aggregation_result.items()}
self.status = self.FINISHED
except:
self.status = self.FAILED
self.save()
def populate(self):
"""
Compute value count using the objects value count parameters.
"""
# Update status
self.status = self.COMPUTING
self.save()
try:
# Compute aggregate result
agg = Aggregator(
layer_dict=self.layer_names,
formula=self.formula,
zoom=self.zoom,
geom=self.aggregationarea.geom,
acres=self.units.lower() == 'acres',
grouping=self.grouping,
)
aggregation_result = agg.value_count()
self.stats_min, self.stats_max, self.stats_avg, self.stats_std = agg.statistics()
# Track cumulative data to be able to generalize stats over
# multiple aggregation areas.
self.stats_cumsum_t0 = agg._stats_t0
self.stats_cumsum_t1 = agg._stats_t1
self.stats_cumsum_t2 = agg._stats_t2
# Convert values to string for storage in hstore
self.value = {k: str(v) for k, v in aggregation_result.items()}
self.status = self.FINISHED
except:
self.status = self.FAILED
self.save()
def test_layer_stats(self):
# Use a legend with simple int expression
agg = Aggregator(
layer_dict={'a': self.rasterlayer.id},
formula='a',
)
# Get original band metadata
meta = self.rasterlayer.rasterlayerbandmetadata_set.first()
# The comparison here is loose, as one is computed on tiles and the other
# other value are the original band values. So there are differences from
# rescaling.
for dat in zip(agg.statistics(), meta.statistics()):
self.assertAlmostEqual(dat[0], dat[1], 1)
def test_full_mask_data(self):
# Override all tiles to be fully masked.
for tile in self.rasterlayer.rastertile_set.all():
tile.rast.bands[0].data([0], shape=(1, 1))
tile.rast.bands[0].nodata_value = 0
tile.save()
# Use a legend with simple int expression
agg = Aggregator(
layer_dict={'a': self.rasterlayer.id},
formula='a',
)
self.assertEqual((None, None, None, None), agg.statistics())
def test_layer_stats(self):
# Use a legend with simple int expression
agg = Aggregator(
layer_dict={'a': self.rasterlayer.id},
formula='a',
)
# Get original band metadata
meta = self.rasterlayer.rasterlayerbandmetadata_set.first()
# The comparison here is loose, as one is computed on tiles and the other
# other value are the original band values. So there are differences from
# rescaling.
for dat in zip(agg.statistics(), meta.statistics()):
self.assertAlmostEqual(dat[0], dat[1], 1)
def test_full_mask_data(self):
# Override all tiles to be fully masked.
for tile in self.rasterlayer.rastertile_set.all():
tile.rast.bands[0].data([0], shape=(1, 1))
tile.rast.bands[0].nodata_value = 0
tile.save()
# Use a legend with simple int expression
agg = Aggregator(
layer_dict={'a': self.rasterlayer.id},
formula='a',
)
self.assertEqual((None, None, None, None), agg.statistics())
