Skip to content

Commit

Permalink
Merge pull request #8927 from gem/infra-fix-deprecation
Browse files Browse the repository at this point in the history 
Assign data types to connectivity dataframes and use `concat` instead of the deprecated `append`
  • Loading branch information
@ptormene
ptormene authored Aug 7, 2023
2 parents 1d8106a + 8c61a35 commit cdb3628
Show file tree
Hide file tree
Showing 3 changed files with 62 additions and 40 deletions.
1 change: 1 addition & 0 deletions openquake/calculators/event_based_damage.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -289,6 +289,7 @@ def post_execute(self, dummy):
if (hasattr(oq, 'infrastructure_connectivity_analysis')
and oq.infrastructure_connectivity_analysis):

logging.info('Running connectivity analysis')
conn_results = connectivity.analysis(self.datastore)
self._store_connectivity_analysis_results(conn_results)

Expand Down
2 changes: 2 additions & 0 deletions openquake/calculators/tests/infrastructure_risk_test.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,8 @@

class InfrastructureRiskTestCase(CalculatorTestCase):

# TODO: we need tests also for event-based

def _check_csv_outputs(self, outputs_list, datastore, testcase,
replace_expected=False):
for output in outputs_list:
Expand Down
99 changes: 59 additions & 40 deletions openquake/risklib/connectivity.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -162,17 +162,18 @@ def analyze_taz_nodes(dstore, exposure_df, G_original, TAZ_nodes, eff_nodes,
sum_connectivity_loss_pcl / eff_inv_time)
avg_connectivity_loss_wcl = sum_connectivity_loss_wcl/eff_inv_time
avg_connectivity_loss_eff = sum_connectivity_loss_eff/eff_inv_time
taz_cl["PCL_node"] /= eff_inv_time
taz_cl["WCL_node"] /= eff_inv_time
node_el["EL"] /= eff_inv_time

elif calculation_mode == "scenario_damage":
num_events = len(damage_df.reset_index().event_id.unique())
avg_connectivity_loss_pcl = sum_connectivity_loss_pcl / num_events
avg_connectivity_loss_wcl = sum_connectivity_loss_wcl / num_events
avg_connectivity_loss_eff = sum_connectivity_loss_eff / num_events
taz_cl.loc[:, "PCL_node"] = taz_cl["PCL_node"].apply(
lambda x: x/num_events)
taz_cl.loc[:, "WCL_node"] = taz_cl["WCL_node"].apply(
lambda x: x/num_events)
node_el.loc[:, "EL"] = node_el["EL"].apply(lambda x: x/num_events)
taz_cl["PCL_node"] /= num_events
taz_cl["WCL_node"] /= num_events
node_el["EL"] /= num_events

taz_cl.drop(columns=['ordinal'], inplace=True)
node_el.drop(columns=['ordinal'], inplace=True)
Expand Down Expand Up @@ -215,21 +216,21 @@ def analyze_demand_nodes(dstore, exposure_df, G_original, eff_nodes,
sum_connectivity_loss_wcl / eff_inv_time)
avg_connectivity_loss_eff = (
sum_connectivity_loss_eff / eff_inv_time)
dem_cl["Isolation_node"] /= eff_inv_time
dem_cl["PCL_node"] /= eff_inv_time
dem_cl["WCL_node"] /= eff_inv_time
node_el["EL"] /= eff_inv_time

elif calculation_mode == "scenario_damage":
num_events = len(damage_df.reset_index().event_id.unique())
avg_connectivity_loss_ccl = sum_connectivity_loss_ccl / num_events
avg_connectivity_loss_pcl = sum_connectivity_loss_pcl / num_events
avg_connectivity_loss_wcl = sum_connectivity_loss_wcl / num_events
avg_connectivity_loss_eff = sum_connectivity_loss_eff/num_events
dem_cl.loc[:, "Isolation_node"] = dem_cl["Isolation_node"].apply(
lambda x: x/num_events)
dem_cl.loc[:, "PCL_node"] = dem_cl["PCL_node"].apply(
lambda x: x/num_events)
dem_cl.loc[:, "WCL_node"] = dem_cl["WCL_node"].apply(
lambda x: x/num_events)
node_el.loc[:, "EL"] = node_el["EL"].apply(
lambda x: x/num_events)
dem_cl["Isolation_node"] /= num_events
dem_cl["PCL_node"] /= num_events
dem_cl["WCL_node"] /= num_events
node_el["EL"] /= num_events

dem_cl.drop(columns=['ordinal'], inplace=True)
node_el.drop(columns=['ordinal'], inplace=True)
Expand Down Expand Up @@ -258,11 +259,12 @@ def analyze_generic_nodes(dstore, exposure_df, G_original, eff_nodes,
num_lt_samples = dstore["oqparam"].number_of_logic_tree_samples
eff_inv_time = inv_time * ses_per_ltp * num_lt_samples
avg_connectivity_loss_eff = sum_connectivity_loss_eff/eff_inv_time
node_el["EL"] /= eff_inv_time

elif calculation_mode == "scenario_damage":
num_events = len(damage_df.reset_index().event_id.unique())
avg_connectivity_loss_eff = sum_connectivity_loss_eff/num_events
node_el.loc[:, "EL"] = node_el["EL"].apply(lambda x: x/num_events)
node_el["EL"] /= num_events

node_el.drop(columns=['ordinal'], inplace=True)

Expand Down Expand Up @@ -425,15 +427,19 @@ def ELWCLPCLCCL_demand(exposure_df, G_original, eff_nodes, demand_nodes,
ccl_table.set_index('id', inplace=True)
pcl_table.set_index('id', inplace=True)
wcl_table.set_index('id', inplace=True)
eff_table.set_index("id", inplace=True)
eff_table.set_index('id', inplace=True)

# Create an empty dataframe with columns "event_id" and
# "CCL"/"PCL"/"WCL"/"EL"

event_connectivity_loss_ccl = pd.DataFrame(columns=['event_id', 'CCL'])
event_connectivity_loss_pcl = pd.DataFrame(columns=['event_id', 'PCL'])
event_connectivity_loss_wcl = pd.DataFrame(columns=['event_id', 'WCL'])
event_connectivity_loss_eff = pd.DataFrame(columns=['event_id', 'EL'])
event_connectivity_loss_ccl = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'CCL': pd.Series(dtype=float)})
event_connectivity_loss_pcl = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'PCL': pd.Series(dtype=float)})
event_connectivity_loss_wcl = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'WCL': pd.Series(dtype=float)})
event_connectivity_loss_eff = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'EL': pd.Series(dtype=float)})

# To check the the values for each node before the earthquake event

Expand Down Expand Up @@ -509,16 +515,21 @@ def ELWCLPCLCCL_demand(exposure_df, G_original, eff_nodes, demand_nodes,
Glo_eff0_per_event - Glo_eff_per_event)/Glo_eff0_per_event

# Storing the value of performance indicators for each event
event_connectivity_loss_ccl = event_connectivity_loss_ccl.append(
{'event_id': event_id, 'CCL': CCL_per_event}, ignore_index=True)
event_connectivity_loss_pcl = event_connectivity_loss_pcl.append(
{'event_id': event_id, 'PCL': PCL_mean_per_event},
event_connectivity_loss_ccl = pd.concat(
[event_connectivity_loss_ccl, pd.DataFrame.from_records(
[{'event_id': event_id, 'CCL': CCL_per_event}])],
ignore_index=True)
event_connectivity_loss_pcl = pd.concat(
[event_connectivity_loss_pcl, pd.DataFrame.from_records(
[{'event_id': event_id, 'PCL': PCL_mean_per_event}])],
ignore_index=True)
event_connectivity_loss_wcl = event_connectivity_loss_wcl.append(
{'event_id': event_id, 'WCL': WCL_mean_per_event},
event_connectivity_loss_wcl = pd.concat(
[event_connectivity_loss_wcl, pd.DataFrame.from_records(
[{'event_id': event_id, 'WCL': WCL_mean_per_event}])],
ignore_index=True)
event_connectivity_loss_eff = event_connectivity_loss_eff.append(
{'event_id': event_id, 'EL': Glo_effloss_per_event},
event_connectivity_loss_eff = pd.concat(
[event_connectivity_loss_eff, pd.DataFrame.from_records(
[{'event_id': event_id, 'EL': Glo_effloss_per_event}])],
ignore_index=True)

# To store the sum of performance indicator at nodal level to calulate
Expand Down Expand Up @@ -566,15 +577,18 @@ def ELWCLPCLloss_TAZ(exposure_df, G_original, TAZ_nodes,
pcl_table = pd.DataFrame({'id': TAZ_nodes})
wcl_table = pd.DataFrame({'id': TAZ_nodes})
eff_table = pd.DataFrame({'id': eff_nodes})
eff_table.set_index("id", inplace=True)
eff_table.set_index('id', inplace=True)
pcl_table.set_index('id', inplace=True)
wcl_table.set_index('id', inplace=True)

# Create an empty dataframe with columns "event_id" and
# "CCL"/"PCL"/"WCL"/"EL"
event_connectivity_loss_pcl = pd.DataFrame(columns=['event_id', 'PCL'])
event_connectivity_loss_wcl = pd.DataFrame(columns=['event_id', 'WCL'])
event_connectivity_loss_eff = pd.DataFrame(columns=['event_id', 'EL'])
event_connectivity_loss_pcl = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'PCL': pd.Series(dtype=float)})
event_connectivity_loss_wcl = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'WCL': pd.Series(dtype=float)})
event_connectivity_loss_eff = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'EL': pd.Series(dtype=float)})

# To check the the values for each node before the earthquake event

Expand Down Expand Up @@ -640,14 +654,17 @@ def ELWCLPCLloss_TAZ(exposure_df, G_original, TAZ_nodes,
Glo_eff0_per_event - Glo_eff_per_event)/Glo_eff0_per_event

# Storing the value of performance indicators for each event
event_connectivity_loss_pcl = event_connectivity_loss_pcl.append(
{'event_id': event_id, 'PCL': PCL_mean_per_event},
event_connectivity_loss_pcl = pd.concat(
[event_connectivity_loss_pcl, pd.DataFrame.from_records(
[{'event_id': event_id, 'PCL': PCL_mean_per_event}])],
ignore_index=True)
event_connectivity_loss_wcl = event_connectivity_loss_wcl.append(
{'event_id': event_id, 'WCL': WCL_mean_per_event},
event_connectivity_loss_wcl = pd.concat(
[event_connectivity_loss_wcl, pd.DataFrame.from_records(
[{'event_id': event_id, 'WCL': WCL_mean_per_event}])],
ignore_index=True)
event_connectivity_loss_eff = event_connectivity_loss_eff.append(
{'event_id': event_id, 'EL': Glo_effloss_per_event},
event_connectivity_loss_eff = pd.concat(
[event_connectivity_loss_eff, pd.DataFrame.from_records(
[{'event_id': event_id, 'EL': Glo_effloss_per_event}])],
ignore_index=True)

# To store the sum of performance indicator at nodal level to calulate
Expand Down Expand Up @@ -684,7 +701,8 @@ def EL_node(exposure_df, G_original, eff_nodes, damage_df, g_type):
eff_table.set_index("id", inplace=True)

# Create an empty dataframe with columns "event_id" and "EL"
event_connectivity_loss_eff = pd.DataFrame(columns=['event_id', 'EL'])
event_connectivity_loss_eff = pd.DataFrame(
{'event_id': pd.Series(dtype=int), 'EL': pd.Series(dtype=float)})

# To check the the values for each node before the earthquake event

Expand Down Expand Up @@ -717,8 +735,9 @@ def EL_node(exposure_df, G_original, eff_nodes, damage_df, g_type):
Glo_eff0_per_event - Glo_eff_per_event)/Glo_eff0_per_event

# Storing the value of performance indicators for each event
event_connectivity_loss_eff = event_connectivity_loss_eff.append(
{'event_id': event_id, 'EL': Glo_effloss_per_event},
event_connectivity_loss_eff = pd.concat(
[event_connectivity_loss_eff, pd.DataFrame.from_records(
[{'event_id': event_id, 'EL': Glo_effloss_per_event}])],
ignore_index=True)

# To store the sum of performance indicator at nodal level to calulate
Expand Down

0 comments on commit cdb3628

Please sign in to comment.