setting = "classic" # classic, steep, or floodplain
# name of the folder of the desired DCEs for the analysis
DCE1_name = "DCE_01"
DCE2_name = 'DCE_02'
# DCE 1 Parameters
DCE1_date = "201108"
DCE1_image_source = 'drone'
DCE1_date_name = "August 2019"
DCE1_flow_stage = 'moderate'
DCE1_active = 'Yes'
DCE1_maintained = 'No'
DCE1_res = '.02'
# DCE 2 Parameters
DCE2_date = '201108'
DCE2_image_source = 'google earth'
DCE2_date_name = "Undammed"
DCE2_flow_stage = 'moderate'
DCE2_active = 'Yes'
DCE2_maintained = 'Yes'
DCE2_res = '.46'
calculate_metrics(project_path, RS_folder_name, DEM, mapper, project_name,
site_name, DCE1_name, DCE1_date, DCE1_image_source,
DCE2_image_source, DCE1_date_name, DCE2_date_name,
DCE1_flow_stage, DCE1_active, DCE1_maintained, DCE2_name,
DCE2_date, DCE2_flow_stage, DCE2_active, DCE2_maintained,
DCE1_res, DCE2_res, setting, huc8)
################################
if p == t:
batch_corrects += 1
if i % print_every == 0 and i > 0:
print('[%d, %5d] loss: %.7f' %
(epoch + 1, i + 1, running_loss / print_every))
print('real: ', str(tmp_target), '----- predicted: ',
str(tmp_predicted))
running_loss = 0.0
print()
total_loss.append((epoch_loss / len(train_data_loader)))
total_corrects += batch_corrects
auc, pr_auc, average_precision, average_recall = calculate_metrics(
targets=targets,
predictions=predictions,
bin_predictions=bin_predictions)
print('\nEpoch %d/%d, Accuracy: %.3f' %
(epoch + 1, num_epochs, batch_corrects / len(train_data_loader)))
print('\nEpoch %d/%d, Loss : %.7f' %
(epoch + 1, num_epochs, (epoch_loss / len(train_data_loader))))
print('\nEpoch %d/%d, AUC: %.3f' % (epoch + 1, num_epochs, auc))
print('\nEpoch %d/%d, PR-AUC: %.3f' % (epoch + 1, num_epochs, pr_auc))
print('\nEpoch %d/%d, Precision: %.3f' %
(epoch + 1, num_epochs, average_precision))
print(
'\nEpoch %d/%d, Recall: %.3f' %
(epoch + 1, num_epochs, average_recall), '\n')
print('Finished Training')
# 2.5 set index for easy concat
dist_df.index = df.index
dist_df.index.set_names(["date"], inplace=True)
# 2.6 add genenric col names
new_col = [
col.replace("{}.".format(district), "")
for col in list(dist_df.columns)
]
dist_df.rename(dict(zip(list(dist_df.columns), new_col)),
axis=1,
inplace=True)
# 2.7 Output to CSV
logging.info("Writing data to output file")
dist_df.to_csv(output_file,
mode="w" if header else "a",
header=header)
# Calculate metrics
logging.info("calculating metrics for {}".format(district))
calculate_metrics.calculate_metrics(
dist_df,
header=header,
hospitalizations=hospitalizations,
output=metrics_file)
header = False
except Exception as e:
logging.exception("Error Occurred")