def test_caching(self):
"""
Test that the caching mechanism works and we can turn it on/off
"""
clear_cache()
self.assertEqual(len(ZonalStatsCache.objects.all()), 0)
zonal = zonal_stats(self.polygons[0], self.rast)
self.assertEqual(zonal.from_cache, False)
self.assertEqual(len(ZonalStatsCache.objects.all()), 1)
zonal = zonal_stats(self.polygons[0], self.rast)
self.assertEqual(zonal.from_cache, True)
self.assertEqual(len(ZonalStatsCache.objects.all()), 1)
zonal = zonal_stats(self.polygons[0], self.rast, read_cache=False)
self.assertEqual(zonal.from_cache, False)
self.assertEqual(len(ZonalStatsCache.objects.all()), 1)
zonal = zonal_stats(self.polygons[3], self.rast)
self.assertEqual(zonal.from_cache, False)
self.assertEqual(len(ZonalStatsCache.objects.all()), 2)
zonal = zonal_stats(self.polygons[3], self.rast)
self.assertEqual(zonal.from_cache, True)
self.assertEqual(len(ZonalStatsCache.objects.all()), 2)
clear_cache()
self.assertEqual(len(ZonalStatsCache.objects.all()), 0)
def test_caching(self):
"""
Test that the caching mechanism works and we can turn it on/off
"""
clear_cache()
self.assertEqual(len(ZonalStatsCache.objects.all()), 0)
zonal = zonal_stats(self.polygons[0], self.rast)
self.assertEqual(zonal.from_cache, False)
self.assertEqual(len(ZonalStatsCache.objects.all()), 1)
zonal = zonal_stats(self.polygons[0], self.rast)
self.assertEqual(zonal.from_cache, True)
self.assertEqual(len(ZonalStatsCache.objects.all()), 1)
zonal = zonal_stats(self.polygons[0], self.rast, read_cache=False)
self.assertEqual(zonal.from_cache, False)
self.assertEqual(len(ZonalStatsCache.objects.all()), 1)
zonal = zonal_stats(self.polygons[3], self.rast)
self.assertEqual(zonal.from_cache, False)
self.assertEqual(len(ZonalStatsCache.objects.all()), 2)
zonal = zonal_stats(self.polygons[3], self.rast)
self.assertEqual(zonal.from_cache, True)
self.assertEqual(len(ZonalStatsCache.objects.all()), 2)
clear_cache()
self.assertEqual(len(ZonalStatsCache.objects.all()), 0)
def get_water_impacts(bioregion):
wi_geom = RasterDataset.objects.get(name='water_impacts')
wi_stats = zonal_stats(bioregion.output_geom, wi_geom)
if wi_stats.pixels:
total_pixels = float(wi_stats.pixels)
categories = wi_stats.categories.all()
impacts_dict = {}
for cat in categories:
impacts_dict[cat.category] = cat.count
perc_impacts = []
for cat, count in impacts_dict.items():
perc_impacts.append((count / total_pixels, cat))
perc_impacts.sort(reverse=True)
final_impacts = []
for impact in perc_impacts:
if impact[1] == 0:
final_impacts.append(('Less Precipitation', impact[0]))
elif impact[1] == 1:
final_impacts.append(
('No Significant Change in Precipitation', impact[0]))
elif impact[1] == 2:
final_impacts.append(('More Precipitation', impact[0]))
return final_impacts
else:
return [default_value]
def hazard_impact(hazard, bioregion):
hzd_gdp_geom = RasterDataset.objects.get(name=hazard + '_impact')
hzd_gdp_stats = zonal_stats(bioregion.output_geom, hzd_gdp_geom)
if hzd_gdp_stats.sum > 0:
return 'High'
else:
return 'Low'
def stats_for_geom(request, raster_name):
# Confirm that we have a valid polygon geometry
if 'geom_txt' in request.REQUEST:
geom_txt = str(request.REQUEST['geom_txt'])
else:
return HttpResponse("Must supply a geom_txt parameter", status=404)
try:
geom = fromstr(geom_txt)
except:
return HttpResponse(
"Must supply a parsable geom_txt parameter (wkt or json)",
status=404)
# Confirm raster with pk exists
try:
raster = RasterDataset.objects.get(name=raster_name)
except:
return HttpResponse("No raster with pk of %s" % pk, status=404)
#TODO check if continuous
zonal = zonal_stats(geom, raster)
zonal.save()
zqs = ZonalStatsCache.objects.filter(pk=zonal.pk)
data = serializers.serialize("json",
zqs,
fields=('avg', 'min', 'max', 'median', 'mode',
'stdev', 'nulls', 'pixels',
'date_modified', 'raster'))
return HttpResponse(data, mimetype='application/json')
def get_human_influence(bioregion):
hii_geom = RasterDataset.objects.get(name='human_influence_index')
hii_stats = zonal_stats(bioregion.output_geom, hii_geom)
if hii_stats.avg:
return hii_stats.avg / 64
else:
return default_value
def hazard_impact(hazard, bioregion):
hzd_gdp_geom = RasterDataset.objects.get(name=hazard + "_impact")
hzd_gdp_stats = zonal_stats(bioregion.output_geom, hzd_gdp_geom)
if hzd_gdp_stats.sum > 0:
return "High"
else:
return "Low"
def hazard_is_likely(hazard, bioregion):
hzd_geom = RasterDataset.objects.get(name=hazard)
hzd_stats = zonal_stats(bioregion.output_geom, hzd_geom)
if hzd_stats.sum > 0:
return True
else:
return False
def get_human_consumption(bioregion):
hc_geom = RasterDataset.objects.get(name='human_consumption')
hc_stats = zonal_stats(bioregion.output_geom, hc_geom)
if hc_stats.avg:
return int(hc_stats.avg)
else:
return default_value
def get_projected_population(bioregion):
pop_geom = RasterDataset.objects.get(name='population_2015')
pop_stats = zonal_stats(bioregion.output_geom, pop_geom)
if pop_stats.sum:
return int(pop_stats.sum)
else:
return 0
def get_soil_suitability(bioregion):
suit_geom = RasterDataset.objects.get(name='soil_suitability')
suit_stats = zonal_stats(bioregion.output_geom, suit_geom)
if suit_stats.avg:
return suit_stats.avg
else:
return default_value
def get_displaced_populations(bioregion):
pop_geom = RasterDataset.objects.get(name="population_2005")
pop_stats = zonal_stats(bioregion.output_geom, pop_geom)
pop_05_sum = pop_stats.sum
if pop_05_sum in [None, 0]:
return (0, 0), (0, 0), (0, 0)
if report_cache_exists(bioregion, "displaced_people_1m"):
displaced_1m = get_report_cache(bioregion, "displaced_people_1m")
else:
sr_objects = SeaRise1m.objects.filter(
geometry__bboverlaps=bioregion.output_geom
) # .aggregate(Collect('geometry'))
displaced_1m = get_displaced(pop_geom, pop_stats, pop_05_sum, sr_objects, bioregion)
create_report_cache(bioregion, dict(displaced_people_1m=displaced_1m))
if report_cache_exists(bioregion, "displaced_people_3m"):
displaced_3m = get_report_cache(bioregion, "displaced_people_3m")
else:
sr_objects = SeaRise3m.objects.filter(
geometry__bboverlaps=bioregion.output_geom
) # .aggregate(Collect('geometry'))
displaced_3m = get_displaced(pop_geom, pop_stats, pop_05_sum, sr_objects, bioregion)
create_report_cache(bioregion, dict(displaced_people_3m=displaced_3m))
if report_cache_exists(bioregion, "displaced_people_6m"):
displaced_6m = get_report_cache(bioregion, "displaced_people_6m")
else:
sr_objects = SeaRise6m.objects.filter(
geometry__bboverlaps=bioregion.output_geom
) # .aggregate(Collect('geometry'))
displaced_6m = get_displaced(pop_geom, pop_stats, pop_05_sum, sr_objects, bioregion)
create_report_cache(bioregion, dict(displaced_people_6m=displaced_6m))
return displaced_1m, displaced_3m, displaced_6m
def get_soil_suitability(bioregion):
suit_geom = RasterDataset.objects.get(name="soil_suitability")
suit_stats = zonal_stats(bioregion.output_geom, suit_geom)
if suit_stats.avg:
return suit_stats.avg
else:
return default_value
def get_human_consumption(bioregion):
hc_geom = RasterDataset.objects.get(name="human_consumption")
hc_stats = zonal_stats(bioregion.output_geom, hc_geom)
if hc_stats.avg:
return int(hc_stats.avg)
else:
return default_value
def get_urban_pop(pop_2005, bioregion):
if report_cache_exists(bioregion, 'urban_population'):
urban_pop, urban_perc = get_report_cache(bioregion, 'urban_population')
return (urban_pop, urban_perc)
else:
pop_geom = RasterDataset.objects.get(name='population_2005')
urban_objects = UrbanExtent.objects.filter(
geometry__bboverlaps=bioregion.output_geom)
urban_pop = 0
for urban_object in urban_objects:
urban_shape = urban_object.geometry
if not urban_shape.valid:
urban_shape = urban_shape.buffer(0)
if urban_shape.intersects(bioregion.output_geom):
urban_overlap = urban_shape.intersection(bioregion.output_geom)
pop_stats = zonal_stats(urban_overlap, pop_geom)
if pop_stats.sum:
urban_pop += pop_stats.sum
if pop_2005 == 0:
urban_perc = 0
else:
urban_perc = urban_pop / pop_2005
create_report_cache(bioregion,
dict(urban_population=(urban_pop, urban_perc)))
return (urban_pop, urban_perc)
def get_projected_population(bioregion):
pop_geom = RasterDataset.objects.get(name='population_2015')
pop_stats = zonal_stats(bioregion.output_geom, pop_geom)
if pop_stats.sum:
return int(pop_stats.sum)
else:
return 0
def get_human_influence(bioregion):
hii_geom = RasterDataset.objects.get(name="human_influence_index")
hii_stats = zonal_stats(bioregion.output_geom, hii_geom)
if hii_stats.avg:
return hii_stats.avg / 64
else:
return default_value
def hazard_is_likely(hazard, bioregion):
hzd_geom = RasterDataset.objects.get(name=hazard)
hzd_stats = zonal_stats(bioregion.output_geom, hzd_geom)
if hzd_stats.sum > 0:
return True
else:
return False
def get_sea_rise_impacts(bioregion):
s1_geom = RasterDataset.objects.get(name="sea_rise_1m")
s3_geom = RasterDataset.objects.get(name="sea_rise_3m")
s6_geom = RasterDataset.objects.get(name="sea_rise_6m")
s1_stats = zonal_stats(bioregion.output_geom, s1_geom)
s3_stats = zonal_stats(bioregion.output_geom, s3_geom)
s6_stats = zonal_stats(bioregion.output_geom, s6_geom)
land_loss_1m = s1_stats.sum * sea_level_grid_size
land_loss_3m = s3_stats.sum * sea_level_grid_size
land_loss_6m = s6_stats.sum * sea_level_grid_size
land_loss_1m_km = int(convert_float_to_area_display_units(land_loss_1m))
land_loss_1m_mi = int(convert_sq_km_to_sq_mi(land_loss_1m_km))
land_loss_3m_km = int(convert_float_to_area_display_units(land_loss_3m))
land_loss_3m_mi = int(convert_sq_km_to_sq_mi(land_loss_3m_km))
land_loss_6m_km = int(convert_float_to_area_display_units(land_loss_6m))
land_loss_6m_mi = int(convert_sq_km_to_sq_mi(land_loss_6m_km))
return (land_loss_1m_km, land_loss_1m_mi), (land_loss_3m_km, land_loss_3m_mi), (land_loss_6m_km, land_loss_6m_mi)
def get_underweight_children(bioregion):
if no_poverty_data(bioregion, 0.1):
return 0
uwc_geom = RasterDataset.objects.get(name="underweight_children")
uwc_stats = zonal_stats(bioregion.output_geom, uwc_geom)
if uwc_stats.sum:
return int(uwc_stats.sum)
else:
return 0
def get_annual_precip(bioregion):
precip_geom = RasterDataset.objects.get(name='annual_precipitation')
precip_stats = zonal_stats(bioregion.output_geom, precip_geom)
if precip_stats.avg:
precip_cm = precip_stats.avg / 10
precip_in = convert_cm_to_in(precip_cm)
return precip_cm, precip_in
else:
return default_value, default_value
def get_annual_temp(bioregion):
temp_geom = RasterDataset.objects.get(name='annual_temperature')
temp_stats = zonal_stats(bioregion.output_geom, temp_geom)
if temp_stats.avg:
temp_c = temp_stats.avg / 10
temp_f = temp_c * 9 / 5. + 32
return temp_c, temp_f
else:
return default_value, default_value
def get_min_temp(bioregion):
min_temp_geom = RasterDataset.objects.get(name='min_temp')
min_temp_stats = zonal_stats(bioregion.output_geom, min_temp_geom)
if min_temp_stats.avg:
min_temp_c = min_temp_stats.avg / 10
min_temp_f = min_temp_c * 9 / 5. + 32
return min_temp_c, min_temp_f
else:
return default_value, default_value
def get_temp_change(bioregion):
tc_geom = RasterDataset.objects.get(name='temp_change')
tc_stats = zonal_stats(bioregion.output_geom, tc_geom)
if tc_stats.avg:
temp_c = tc_stats.avg
temp_f = temp_c * 1.8
return (temp_c, temp_f)
else:
return default_value
def get_min_temp(bioregion):
min_temp_geom = RasterDataset.objects.get(name='min_temp')
min_temp_stats = zonal_stats(bioregion.output_geom, min_temp_geom)
if min_temp_stats.avg:
min_temp_c = min_temp_stats.avg / 10
min_temp_f = min_temp_c * 9 / 5. + 32
return min_temp_c, min_temp_f
else:
return default_value, default_value
def get_annual_precip(bioregion):
precip_geom = RasterDataset.objects.get(name='annual_precipitation')
precip_stats = zonal_stats(bioregion.output_geom, precip_geom)
if precip_stats.avg:
precip_cm = precip_stats.avg / 10
precip_in = convert_cm_to_in(precip_cm)
return precip_cm, precip_in
else:
return default_value, default_value
def get_underweight_children(bioregion):
if no_poverty_data(bioregion, .1):
return 0
uwc_geom = RasterDataset.objects.get(name='underweight_children')
uwc_stats = zonal_stats(bioregion.output_geom, uwc_geom)
if uwc_stats.sum:
return int(uwc_stats.sum)
else:
return 0
def get_temp_change(bioregion):
tc_geom = RasterDataset.objects.get(name="temp_change")
tc_stats = zonal_stats(bioregion.output_geom, tc_geom)
if tc_stats.avg:
temp_c = tc_stats.avg
temp_f = temp_c * 1.8
return (temp_c, temp_f)
else:
return default_value
def get_annual_temp(bioregion):
temp_geom = RasterDataset.objects.get(name='annual_temperature')
temp_stats = zonal_stats(bioregion.output_geom, temp_geom)
if temp_stats.avg:
temp_c = temp_stats.avg / 10
temp_f = temp_c * 9 / 5. + 32
return temp_c, temp_f
else:
return default_value, default_value
def test_webservice(self):
data = {'geom_txt': self.polygons[0].wkt}
#self.settings_manager.set(ROOT_URLCONF = 'madrona.raster_stats.urls')
response = self.client.get('/test_impact/', data)
self.failUnlessEqual(response.status_code, 200)
for obj in serializers.deserialize("json", response.content):
web_zonal = obj.object
util_zonal = zonal_stats(self.polygons[0], self.rast, read_cache=False)
self.failUnlessEqual(web_zonal.avg, util_zonal.avg)
def get_terr_npp_avg(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return 0
npp_geom = RasterDataset.objects.get(name='npp_terr')
npp_stats = zonal_stats(terra_geom, npp_geom)
if npp_stats.avg:
npp_avg = npp_stats.avg / (26064.03459**2)
else:
npp_avg = 0
return npp_avg
def get_sea_rise_impacts(bioregion):
s1_geom = RasterDataset.objects.get(name='sea_rise_1m')
s3_geom = RasterDataset.objects.get(name='sea_rise_3m')
s6_geom = RasterDataset.objects.get(name='sea_rise_6m')
s1_stats = zonal_stats(bioregion.output_geom, s1_geom)
s3_stats = zonal_stats(bioregion.output_geom, s3_geom)
s6_stats = zonal_stats(bioregion.output_geom, s6_geom)
land_loss_1m = s1_stats.sum * sea_level_grid_size
land_loss_3m = s3_stats.sum * sea_level_grid_size
land_loss_6m = s6_stats.sum * sea_level_grid_size
land_loss_1m_km = int(convert_float_to_area_display_units(land_loss_1m))
land_loss_1m_mi = int(convert_sq_km_to_sq_mi(land_loss_1m_km))
land_loss_3m_km = int(convert_float_to_area_display_units(land_loss_3m))
land_loss_3m_mi = int(convert_sq_km_to_sq_mi(land_loss_3m_km))
land_loss_6m_km = int(convert_float_to_area_display_units(land_loss_6m))
land_loss_6m_mi = int(convert_sq_km_to_sq_mi(land_loss_6m_km))
return (land_loss_1m_km,
land_loss_1m_mi), (land_loss_3m_km,
land_loss_3m_mi), (land_loss_6m_km,
land_loss_6m_mi)
def get_terr_npp_avg(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return 0
npp_geom = RasterDataset.objects.get(name='npp_terr')
npp_stats = zonal_stats(terra_geom, npp_geom)
if npp_stats.avg:
npp_avg = npp_stats.avg / (26064.03459 ** 2)
else:
npp_avg = 0
return npp_avg
def test_zonal_util(self):
"""
Tests that starspan works and stuff
"""
# shouldnt have any nulls
zonal = zonal_stats(self.polygons[0], self.rast)
self.assertEqual(zonal.nulls, 0)
# doesnt even touch the raster, all should be null
zonal = zonal_stats(self.polygons[1], self.rast)
self.assertEqual(zonal.pixels, None)
# Partly on and partly off the raster
# no nulls but pixel count should be low
zonal = zonal_stats(self.polygons[2], self.rast)
self.assertEqual(zonal.nulls, 0)
self.assertEqual(zonal.pixels, 225)
# All on the raster but should have nulls
zonal = zonal_stats(self.polygons[3], self.rast)
self.assertEqual(zonal.nulls, 279)
def test_zonal_util(self):
"""
Tests that starspan works and stuff
"""
# shouldnt have any nulls
zonal = zonal_stats(self.polygons[0], self.rast)
self.assertEqual(zonal.nulls,0)
# doesnt even touch the raster, all should be null
zonal = zonal_stats(self.polygons[1], self.rast)
self.assertEqual(zonal.pixels,None)
# Partly on and partly off the raster
# no nulls but pixel count should be low
zonal = zonal_stats(self.polygons[2], self.rast)
self.assertEqual(zonal.nulls,0)
self.assertEqual(zonal.pixels,225)
# All on the raster but should have nulls
zonal = zonal_stats(self.polygons[3], self.rast)
self.assertEqual(zonal.nulls,279)
def test_webservice(self):
data = {'geom_txt': self.polygons[0].wkt}
#self.settings_manager.set(ROOT_URLCONF = 'madrona.raster_stats.urls')
response = self.client.get('/test_impact/', data)
self.failUnlessEqual(response.status_code, 200)
for obj in serializers.deserialize("json", response.content):
web_zonal = obj.object
util_zonal = zonal_stats(self.polygons[0],
self.rast,
read_cache=False)
self.failUnlessEqual(web_zonal.avg, util_zonal.avg)
def get_climate_change_vulnerability(bioregion):
cc_geom = RasterDataset.objects.get(name='climate_impact')
cc_stats = zonal_stats(bioregion.output_geom, cc_geom)
if cc_stats.avg:
if cc_stats.avg < -12: index = 'Low'
elif cc_stats.avg < -4: index = 'Moderately Low'
elif cc_stats.avg < 4: index = 'Moderate'
elif cc_stats.avg < 12: index = 'Moderately High'
else: index = 'High'
return (index, cc_stats.avg)
else:
return default_value
def get_ocn_npp_avg(bioregion):
oceanic_geom = get_oceanic_geom(bioregion)
if oceanic_geom.area == 0:
npp_avg = 0
else:
npp_geom = RasterDataset.objects.get(name='npp_ocn')
npp_stats = zonal_stats(oceanic_geom, npp_geom)
#settling on the following to account for lack of overlap
#(see if statement above which helps with this issue as well)
try:
npp_avg = npp_stats.avg * 365 / 1000 #mg per day converted to g per year
except:
npp_avg = 0
return npp_avg
def get_soil_suitability(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return (0, 0, 0)
suit_geom = RasterDataset.objects.get(name='soil_suitability')
suit_stats = zonal_stats(terra_geom, suit_geom)
if suit_stats.avg:
proportion = suit_stats.avg
area_km, area_mi = get_size(terra_geom)
prop_area_km = area_km * proportion
prop_area_mi = area_mi * proportion
return (proportion, prop_area_km, prop_area_mi)
else:
return (0, 0, 0)
def get_ocn_npp_avg(bioregion):
oceanic_geom = get_oceanic_geom(bioregion)
if oceanic_geom.area == 0:
npp_avg = 0
else:
npp_geom = RasterDataset.objects.get(name='npp_ocn')
npp_stats = zonal_stats(oceanic_geom, npp_geom)
#settling on the following to account for lack of overlap
#(see if statement above which helps with this issue as well)
try:
npp_avg = npp_stats.avg * 365 / 1000 #mg per day converted to g per year
except:
npp_avg = 0
return npp_avg
def get_soil_suitability(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return (0, 0, 0)
suit_geom = RasterDataset.objects.get(name='soil_suitability')
suit_stats = zonal_stats(terra_geom, suit_geom)
if suit_stats.avg:
proportion = suit_stats.avg
area_km, area_mi = get_size(terra_geom)
prop_area_km = area_km * proportion
prop_area_mi = area_mi * proportion
return (proportion, prop_area_km, prop_area_mi)
else:
return (0, 0, 0)
def get_displaced(pop_geom, pop_stats, pop_05_sum, sr_objects, bioregion):
sr_geoms = [
sr.geometry.intersection(bioregion.output_geom) for sr in sr_objects
if sr.geometry.intersects(bioregion.output_geom)
]
sum_pop = 0
for sr_geom in sr_geoms:
if not sr_geom.valid:
sr_geom = sr_geom.buffer(0)
pop_stats = zonal_stats(sr_geom, pop_geom)
if pop_stats.sum:
sum_pop += pop_stats.sum
perc_pop = sum_pop / pop_05_sum
sum_pop = int(sum_pop / 1000) * 1000
return (sum_pop, perc_pop)
def get_water_stress(bioregion):
ws_geom = RasterDataset.objects.get(name='water_stress')
ws_stats = zonal_stats(bioregion.output_geom, ws_geom)
if ws_stats.avg:
rwsi = ws_stats.avg
if rwsi >= 3.5:
rating = 'High'
elif rwsi >= 2.5:
rating = 'Moderate'
elif rwsi >= 1.5:
rating = 'Low'
else:
rating = 'Unstressed'
return rwsi, rating
return default_value, default_value
def get_proportion_equipped_for_irrigation(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return (0, 0, 0)
irrig_geom = RasterDataset.objects.get(name='irrig_equipped')
irrig_stats = zonal_stats(terra_geom, irrig_geom)
if irrig_stats.sum:
hectares = irrig_stats.sum
area_km, area_mi = get_size(terra_geom)
proportion = hectares / 100 / area_km
prop_area_km = area_km * proportion
prop_area_mi = area_mi * proportion
return (proportion, prop_area_km, prop_area_mi)
else:
return (0, 0, 0)
def get_water_stress(bioregion):
ws_geom = RasterDataset.objects.get(name="water_stress")
ws_stats = zonal_stats(bioregion.output_geom, ws_geom)
if ws_stats.avg:
rwsi = ws_stats.avg
if rwsi >= 3.5:
rating = "High"
elif rwsi >= 2.5:
rating = "Moderate"
elif rwsi >= 1.5:
rating = "Low"
else:
rating = "Unstressed"
return rwsi, rating
return default_value, default_value
def get_proportion_equipped_for_irrigation(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return (0, 0, 0)
irrig_geom = RasterDataset.objects.get(name='irrig_equipped')
irrig_stats = zonal_stats(terra_geom, irrig_geom)
if irrig_stats.sum:
hectares = irrig_stats.sum
area_km, area_mi = get_size(terra_geom)
proportion = hectares / 100 / area_km
prop_area_km = area_km * proportion
prop_area_mi = area_mi * proportion
return (proportion, prop_area_km, prop_area_mi)
else:
return (0, 0, 0)
def get_displaced(pop_geom, pop_stats, pop_05_sum, sr_objects, bioregion):
sr_geoms = [
sr.geometry.intersection(bioregion.output_geom)
for sr in sr_objects
if sr.geometry.intersects(bioregion.output_geom)
]
sum_pop = 0
for sr_geom in sr_geoms:
if not sr_geom.valid:
sr_geom = sr_geom.buffer(0)
pop_stats = zonal_stats(sr_geom, pop_geom)
if pop_stats.sum:
sum_pop += pop_stats.sum
perc_pop = sum_pop / pop_05_sum
sum_pop = int(sum_pop / 1000) * 1000
return (sum_pop, perc_pop)
def get_climate_change_vulnerability(bioregion):
cc_geom = RasterDataset.objects.get(name="climate_impact")
cc_stats = zonal_stats(bioregion.output_geom, cc_geom)
if cc_stats.avg:
if cc_stats.avg < -12:
index = "Low"
elif cc_stats.avg < -4:
index = "Moderately Low"
elif cc_stats.avg < 4:
index = "Moderate"
elif cc_stats.avg < 12:
index = "Moderately High"
else:
index = "High"
return (index, cc_stats.avg)
else:
return default_value
def get_agricultural_impacts(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return [default_value]
ag_geom = RasterDataset.objects.get(name='ag_impacts')
ag_stats = zonal_stats(terra_geom, ag_geom)
if ag_stats.pixels:
total_pixels = float(ag_stats.pixels)
categories = ag_stats.categories.all()
impacts_dict = {}
for cat in categories:
impacts_dict[cat.category] = cat.count
perc_impacts = []
for cat, count in impacts_dict.items():
perc_impacts.append((count / total_pixels, cat))
perc_impacts.sort(reverse=True)
final_impacts = []
for impact in perc_impacts:
if impact[1] == 0:
final_impacts.append(('No Data', impact[0]))
if impact[0] > .9:
final_impacts = [default_value]
break
elif impact[1] == 1:
final_impacts.append(
('a Major Increase in Production', impact[0]))
elif impact[1] == 2:
final_impacts.append(
('an General Increase in Production', impact[0]))
elif impact[1] == 3:
final_impacts.append(
('a Slight Increase in Production', impact[0]))
elif impact[1] == 4:
final_impacts.append(
('a Slight Decrease in Production', impact[0]))
elif impact[1] == 5:
final_impacts.append(
('a General Decrease in Production', impact[0]))
elif impact[1] == 6:
final_impacts.append(
('a Major Decrease in Production', impact[0]))
return final_impacts
else:
return [default_value]
def get_marine_ecosystem_impacts(bioregion):
oceanic_geom = get_oceanic_geom(bioregion)
if oceanic_geom.area == 0:
return 0
mi_geom = RasterDataset.objects.get(name='ocean_impact')
mi_stats = zonal_stats(oceanic_geom, mi_geom)
if mi_stats.avg:
avg = round(mi_stats.avg, 1)
if avg < 1.4:
impact = 'Very Low'
elif avg < 4.95:
impact = 'Low'
elif avg < 8.47:
impact = 'Medium'
elif avg < 12:
impact = 'Medium High'
elif avg < 15.52:
impact = 'High'
else:
impact = 'Very High'
return impact
return default_value
def get_marine_ecosystem_impacts(bioregion):
oceanic_geom = get_oceanic_geom(bioregion)
if oceanic_geom.area == 0:
return 0
mi_geom = RasterDataset.objects.get(name="ocean_impact")
mi_stats = zonal_stats(oceanic_geom, mi_geom)
if mi_stats.avg:
avg = round(mi_stats.avg, 1)
if avg < 1.4:
impact = "Very Low"
elif avg < 4.95:
impact = "Low"
elif avg < 8.47:
impact = "Medium"
elif avg < 12:
impact = "Medium High"
elif avg < 15.52:
impact = "High"
else:
impact = "Very High"
return impact
return default_value
def get_agricultural_impacts(bioregion):
terra_geom = get_terra_geom(bioregion)
if terra_geom.area == 0.0:
return [default_value]
ag_geom = RasterDataset.objects.get(name="ag_impacts")
ag_stats = zonal_stats(terra_geom, ag_geom)
if ag_stats.pixels:
total_pixels = float(ag_stats.pixels)
categories = ag_stats.categories.all()
impacts_dict = {}
for cat in categories:
impacts_dict[cat.category] = cat.count
perc_impacts = []
for cat, count in impacts_dict.items():
perc_impacts.append((count / total_pixels, cat))
perc_impacts.sort(reverse=True)
final_impacts = []
for impact in perc_impacts:
if impact[1] == 0:
final_impacts.append(("No Data", impact[0]))
if impact[0] > 0.9:
final_impacts = [default_value]
break
elif impact[1] == 1:
final_impacts.append(("a Major Increase in Production", impact[0]))
elif impact[1] == 2:
final_impacts.append(("an General Increase in Production", impact[0]))
elif impact[1] == 3:
final_impacts.append(("a Slight Increase in Production", impact[0]))
elif impact[1] == 4:
final_impacts.append(("a Slight Decrease in Production", impact[0]))
elif impact[1] == 5:
final_impacts.append(("a General Decrease in Production", impact[0]))
elif impact[1] == 6:
final_impacts.append(("a Major Decrease in Production", impact[0]))
return final_impacts
else:
return [default_value]
def get_displaced_populations(bioregion):
pop_geom = RasterDataset.objects.get(name='population_2005')
pop_stats = zonal_stats(bioregion.output_geom, pop_geom)
pop_05_sum = pop_stats.sum
if pop_05_sum in [None, 0]:
return (0, 0), (0, 0), (0, 0)
if report_cache_exists(bioregion, 'displaced_people_1m'):
displaced_1m = get_report_cache(bioregion, 'displaced_people_1m')
else:
sr_objects = SeaRise1m.objects.filter(
geometry__bboverlaps=bioregion.output_geom
) #.aggregate(Collect('geometry'))
displaced_1m = get_displaced(pop_geom, pop_stats, pop_05_sum,
sr_objects, bioregion)
create_report_cache(bioregion, dict(displaced_people_1m=displaced_1m))
if report_cache_exists(bioregion, 'displaced_people_3m'):
displaced_3m = get_report_cache(bioregion, 'displaced_people_3m')
else:
sr_objects = SeaRise3m.objects.filter(
geometry__bboverlaps=bioregion.output_geom
) #.aggregate(Collect('geometry'))
displaced_3m = get_displaced(pop_geom, pop_stats, pop_05_sum,
sr_objects, bioregion)
create_report_cache(bioregion, dict(displaced_people_3m=displaced_3m))
if report_cache_exists(bioregion, 'displaced_people_6m'):
displaced_6m = get_report_cache(bioregion, 'displaced_people_6m')
else:
sr_objects = SeaRise6m.objects.filter(
geometry__bboverlaps=bioregion.output_geom
) #.aggregate(Collect('geometry'))
displaced_6m = get_displaced(pop_geom, pop_stats, pop_05_sum,
sr_objects, bioregion)
create_report_cache(bioregion, dict(displaced_people_6m=displaced_6m))
return displaced_1m, displaced_3m, displaced_6m
def get_urban_pop(pop_2005, bioregion):
if report_cache_exists(bioregion, 'urban_population'):
urban_pop, urban_perc = get_report_cache(bioregion, 'urban_population')
return (urban_pop, urban_perc)
else:
pop_geom = RasterDataset.objects.get(name='population_2005')
urban_objects = UrbanExtent.objects.filter(geometry__bboverlaps=bioregion.output_geom)
urban_pop = 0
for urban_object in urban_objects:
urban_shape = urban_object.geometry
if not urban_shape.valid:
urban_shape = urban_shape.buffer(0)
if urban_shape.intersects(bioregion.output_geom):
urban_overlap = urban_shape.intersection(bioregion.output_geom)
pop_stats = zonal_stats(urban_overlap, pop_geom)
if pop_stats.sum:
urban_pop += pop_stats.sum
if pop_2005 == 0:
urban_perc = 0
else:
urban_perc = urban_pop / pop_2005
create_report_cache(bioregion, dict(urban_population=(urban_pop, urban_perc)))
return (urban_pop, urban_perc)
def stats_for_geom(request, raster_name):
# Confirm that we have a valid polygon geometry
if 'geom_txt' in request.REQUEST:
geom_txt = str(request.REQUEST['geom_txt'])
else:
return HttpResponse("Must supply a geom_txt parameter", status=404)
try:
geom = fromstr(geom_txt)
except:
return HttpResponse("Must supply a parsable geom_txt parameter (wkt or json)", status=404)
# Confirm raster with pk exists
try:
raster = RasterDataset.objects.get(name=raster_name)
except:
return HttpResponse("No raster with pk of %s" % pk, status=404)
#TODO check if continuous
zonal = zonal_stats(geom, raster)
zonal.save()
zqs = ZonalStatsCache.objects.filter(pk=zonal.pk)
data = serializers.serialize("json", zqs, fields=('avg','min','max','median','mode','stdev','nulls','pixels','date_modified','raster'))
return HttpResponse(data, mimetype='application/json')
def get_water_impacts(bioregion):
wi_geom = RasterDataset.objects.get(name="water_impacts")
wi_stats = zonal_stats(bioregion.output_geom, wi_geom)
if wi_stats.pixels:
total_pixels = float(wi_stats.pixels)
categories = wi_stats.categories.all()
impacts_dict = {}
for cat in categories:
impacts_dict[cat.category] = cat.count
perc_impacts = []
for cat, count in impacts_dict.items():
perc_impacts.append((count / total_pixels, cat))
perc_impacts.sort(reverse=True)
final_impacts = []
for impact in perc_impacts:
if impact[1] == 0:
final_impacts.append(("Less Precipitation", impact[0]))
elif impact[1] == 1:
final_impacts.append(("No Significant Change in Precipitation", impact[0]))
elif impact[1] == 2:
final_impacts.append(("More Precipitation", impact[0]))
return final_impacts
else:
return [default_value]
def test_categories(self):
zonal = zonal_stats(self.polygons[0], self.rast)
sumpix = 0
for zc in zonal.categories.all():
sumpix += zc.count
self.assertEqual(zonal.pixels, sumpix)
def get_infant_deaths(bioregion):
id_geom = RasterDataset.objects.get(name="infant_deaths")
id_stats = zonal_stats(bioregion.output_geom, id_geom)
return int(id_stats.sum)
def get_languages(bioregion):
if report_cache_exists(bioregion, 'languages'):
language_names = get_report_cache(bioregion, 'languages')
return language_names
else:
pop_geom = RasterDataset.objects.get(name='population_2005')
languages = Language.objects.filter(
geometry__bboverlaps=bioregion.output_geom)
language_dict = {}
pop_total = 0
for language in languages:
#noticing issue on ninkasi and dionysus (not on local machine)
#dionysus is using geos 3.1.0
#local machine (which does not crash when intersecting) is using 3.2.2
# select postgis_full_version(); gave me geos version
#possible solutions could be to update geos (not really an option) or perhaps django has an update that catches that error...
#the following error is output on dionysus when attempting intersection (after buffer(0)) on some problem geoemetries
"""
GEOS_NOTICE: Self-intersection at or near point 1.36725e+07 -289750
bufferOriginalPrecision failed (TopologyException: unable to assign hole to a shell), trying with reduced precision
recomputing with precision scale factor = 10000
Scaler: offsetX,Y: 0,0 scaleFactor: 10000
ReScaler: offsetX,Y: 0,0 scaleFactor: 10000
recomputing with precision scale factor = 1000
Scaler: offsetX,Y: 0,0 scaleFactor: 1000
ReScaler: offsetX,Y: 0,0 scaleFactor: 1000
recomputing with precision scale factor = 100
Scaler: offsetX,Y: 0,0 scaleFactor: 100
ReScaler: offsetX,Y: 0,0 scaleFactor: 100
recomputing with precision scale factor = 10
Scaler: offsetX,Y: 0,0 scaleFactor: 10
python: ../../source/headers/geos/noding/SegmentString.h:175: void geos::noding::SegmentString::testInvariant() const: Assertion `pts->size() > 1' failed.
"""
try:
if language.geometry.valid:
language_intersection = language.geometry.intersection(
bioregion.output_geom)
else:
language_intersection = language.geometry.buffer(
0).intersection(bioregion.output_geom)
area = language_intersection.area
except:
area = 0
if area > 0:
if language.name_prop is None:
name = 'No Data'
else:
#name = (language.nam_ansi, language.familyprop)
name = language.name_prop
#area = geometry_area_in_display_units(language.geometry.intersection(bioregion.output_geom))
if language_intersection.valid:
pop_stats = zonal_stats(language_intersection, pop_geom)
else:
buffered_intersection = language_intersection.buffer(0)
pop_stats = zonal_stats(buffered_intersection, pop_geom)
if pop_stats and pop_stats.sum:
pop = pop_stats.sum
else:
pop = 0
pop_total += pop
if name in language_dict.keys():
language_dict[name] += pop
else:
language_dict[name] = pop
expected_pop = get_population(bioregion)
if expected_pop > pop_total:
language_dict['No Data'] = expected_pop - pop_total
language_tuples = [(pop, name) for name, pop in language_dict.items()]
language_tuples.sort(reverse=True)
#language_names = [name for (pop, name) in language_tuples]
language_names = language_tuples
create_report_cache(bioregion, dict(languages=language_names))
return language_names
def get_languages(bioregion):
if report_cache_exists(bioregion, 'languages'):
language_names = get_report_cache(bioregion, 'languages')
return language_names
else:
pop_geom = RasterDataset.objects.get(name='population_2005')
languages = Language.objects.filter(geometry__bboverlaps=bioregion.output_geom)
language_dict = {}
pop_total = 0
for language in languages:
#noticing issue on ninkasi and dionysus (not on local machine)
#dionysus is using geos 3.1.0
#local machine (which does not crash when intersecting) is using 3.2.2
# select postgis_full_version(); gave me geos version
#possible solutions could be to update geos (not really an option) or perhaps django has an update that catches that error...
#the following error is output on dionysus when attempting intersection (after buffer(0)) on some problem geoemetries
"""
GEOS_NOTICE: Self-intersection at or near point 1.36725e+07 -289750
bufferOriginalPrecision failed (TopologyException: unable to assign hole to a shell), trying with reduced precision
recomputing with precision scale factor = 10000
Scaler: offsetX,Y: 0,0 scaleFactor: 10000
ReScaler: offsetX,Y: 0,0 scaleFactor: 10000
recomputing with precision scale factor = 1000
Scaler: offsetX,Y: 0,0 scaleFactor: 1000
ReScaler: offsetX,Y: 0,0 scaleFactor: 1000
recomputing with precision scale factor = 100
Scaler: offsetX,Y: 0,0 scaleFactor: 100
ReScaler: offsetX,Y: 0,0 scaleFactor: 100
recomputing with precision scale factor = 10
Scaler: offsetX,Y: 0,0 scaleFactor: 10
python: ../../source/headers/geos/noding/SegmentString.h:175: void geos::noding::SegmentString::testInvariant() const: Assertion `pts->size() > 1' failed.
"""
try:
if language.geometry.valid:
language_intersection = language.geometry.intersection(bioregion.output_geom)
else:
language_intersection = language.geometry.buffer(0).intersection(bioregion.output_geom)
area = language_intersection.area
except:
area = 0
if area > 0:
if language.name_prop is None:
name = 'No Data'
else:
#name = (language.nam_ansi, language.familyprop)
name = language.name_prop
#area = geometry_area_in_display_units(language.geometry.intersection(bioregion.output_geom))
if language_intersection.valid:
pop_stats = zonal_stats(language_intersection, pop_geom)
else:
buffered_intersection = language_intersection.buffer(0)
pop_stats = zonal_stats(buffered_intersection, pop_geom)
if pop_stats and pop_stats.sum:
pop = pop_stats.sum
else:
pop = 0
pop_total += pop
if name in language_dict.keys():
language_dict[name] += pop
else:
language_dict[name] = pop
expected_pop = get_population(bioregion)
if expected_pop > pop_total:
language_dict['No Data'] = expected_pop - pop_total
language_tuples = [(pop, name) for name,pop in language_dict.items()]
language_tuples.sort(reverse=True)
#language_names = [name for (pop, name) in language_tuples]
language_names = language_tuples
create_report_cache(bioregion, dict(languages=language_names))
return language_names
