the areal_data dataframe with average rainfall data in each areal.
'''
areal_data_copy = areal_data.copy()
points = point_data['geometry']
areas = areal_data['geometry']
results = []
for area in areas:
filtered_points = point_data[points.within(area)]
results.append(filtered_points[column_name].mean())
areal_data_copy = areal_data_copy.assign(**{column_name: results})
return areal_data_copy
if __name__ == '__main__':
rainfall_daily = load_rainfall_data('daily')
flood_daily = load_flood_data('daily')
rain_geo_pandas = convert_to_geopanda(rainfall_daily)
flood_geo_pandas = convert_to_geopanda(flood_daily)
huc8_units_file = '/Users/haigangliu/SpatialTemporalBayes/data/shape_file/hydrologic_HUC8_units/wbdhu8_a_sc.shp'
huc8_units = gpd.read_file(huc8_units_file)
areal_data = fill_area_with_point_data(huc8_units, rain_geo_pandas, 'PRCP')
areal_data = fill_area_with_point_data(areal_data, flood_geo_pandas,
'GAGE_MAX')
test_dataset = areal_data[['NAME', 'geometry', 'PRCP', 'GAGE_MAX']]
self.predictions = np.exp(
np.median(self.simulated_values['y_pred'], axis=0))
l1_loss = np.mean(np.abs(self.predictions - self.y_test))
l2_loss = np.mean(np.square(self.predictions - self.y_test))
self.summary = {'l1_loss': l1_loss, 'l2_loss': l2_loss}
output_df = self.test_loc_cache.copy()
output_df['PRED'] = self.predictions
return self.predictions
if __name__ == '__main__':
from SampleDataLoader import load_rainfall_data
data = load_rainfall_data('monthly')
gp_spatial_model = GPModelSpatial(data,
split_ratio=0.7,
response_var='PRCP')
gp_spatial_model.fit(traceplot_name='test_traceplot.png',
fast_sampling=True)
vars_ = gp_spatial_model.predict()
import pickle
with open('result.pickle', 'wb') as handler:
pickle.dump(gp_spatial_model,
handler,
protocol=pickle.HIGHEST_PROTOCOL)
print(gp_spatial_model.summary)
snapshots = []
indices = self._find_time_index()
start_year_ = self.start_year
start_month_ = self.start_month
start_date_ = self.start_date
for index in indices:
self.sst.data[index,:,:][~self.nc_mask] = np.nan
df_single_month = self._monthly_look_up(index)
df_single_month.loc[:,'MONTH'] = start_month_
df_single_month.loc[:,'YEAR'] = start_year_
snapshots.append(df_single_month)
start_date_ = start_date_ + relativedelta(months=1)
current_month = start_date_.month
current_year = start_date_.year
return pd.concat(snapshots, ignore_index=True)
if __name__ == '__main__':
from SampleDataLoader import load_rainfall_data
new_df = load_rainfall_data()
locs = new_df.groupby('STATION').first().reset_index()[['STATION','LONGITUDE', 'LATITUDE']]
lats_and_lons = locs[['LONGITUDE', 'LATITUDE']]
test_case = SSTcalculator(lats_and_lons, 2011, 10, 2011, 11).run()
print(test_case.SST.describe())