Support historical scenario for score-based risk measures. (#372)

* Support historical scenario for score-based risk measures. Signed-off-by: Joe Moorhouse <[email protected]> * Type fixes Signed-off-by: Joe Moorhouse <[email protected]> --------- Signed-off-by: Joe Moorhouse <[email protected]>
os-climate · Dec 18, 2024 · aac8619 · aac8619
1 parent 7be265b
commit aac8619
Show file tree

Hide file tree

Showing 7 changed files with 40 additions and 36 deletions.
diff --git a/pyproject.toml b/pyproject.toml
@@ -2,7 +2,7 @@
 name = "physrisk-lib"
 # Could test changing the below to be sourced "dynamically"
 # dynamic = ['version']
-version = "0.41.0"
+version = "0.42.0"
 description = "OS-Climate Physical Risk Library"
 authors = [
     {name = "Joe Moorhouse",email = "[email protected]"},

diff --git a/src/physrisk/api/v1/impact_req_resp.py b/src/physrisk/api/v1/impact_req_resp.py
@@ -57,7 +57,7 @@ class AssetImpactRequest(BaseModel):
     # to be deprecated
     scenario: str = Field("rcp8p5", description="Name of scenario ('rcp8p5')")
     year: int = Field(
-        [2050],
+        2050,
         description="""Projection years (e.g. 2030, 2050, 2080). Any year before 2030,
         e.g. 1980, is treated as historical.""",
     )
@@ -72,7 +72,7 @@ class Category(int, Enum):
 
 
 class RiskMeasureDefinition(BaseModel):
-    measure_id: str = Field(None, description="Identifier for the risk measure.")
+    measure_id: str = Field("", description="Identifier for the risk measure.")
     label: str = Field(
         "<short description of the measure, e.g. fractional loss for 1-in-100 year event>",
         description="Short label for the measure quantity.",
@@ -85,7 +85,7 @@ class RiskMeasureDefinition(BaseModel):
 
 class RiskScoreValue(BaseModel):
     value: Category = Field(
-        "", description="Value of the score: red, amber, green, nodata."
+        Category.NODATA, description="Value of the score: red, amber, green, nodata."
     )
     label: str = Field(
         "",
@@ -128,7 +128,7 @@ class RiskMeasureKey(BaseModel):
 
 class RiskMeasuresForAssets(BaseModel):
     key: RiskMeasureKey
-    scores: List[int] = Field(None, description="Identifier for the risk measure.")
+    scores: List[int] = Field([0], description="Identifier for the risk measure.")
     measures_0: List[float]
     measures_1: Optional[List[float]]
 
@@ -156,14 +156,14 @@ class AcuteHazardCalculationDetails(BaseModel):
     hazard_distribution: Distribution
     vulnerability_distribution: VulnerabilityDistrib
     hazard_path: List[str] = Field(
-        "unknown", description="Path to the hazard indicator data source."
+        ["unknown"], description="Path to the hazard indicator data source."
     )
 
 
 class ImpactKey(BaseModel):
     hazard_type: str = Field("", description="Type of the hazard.")
-    scenario_id: str = Field(None, description="Identifier of the scenario.")
-    year: str = Field(None, description="Year of impact.")
+    scenario_id: str = Field("", description="Identifier of the scenario.")
+    year: str = Field("", description="Year of impact.")
 
 
 class AssetSingleImpact(BaseModel):

diff --git a/src/physrisk/data/pregenerated_hazard_model.py b/src/physrisk/data/pregenerated_hazard_model.py
@@ -138,19 +138,21 @@ def _get_hazard_data_batch(
                     )
         except Exception as err:
             # e.g. the requested data is unavailable
-            for _i, req in enumerate(batch):
+            for _, req in enumerate(batch):
                 failed_response = HazardDataFailedResponse(err)
                 responses[req] = failed_response
                 failures.append(failed_response)
 
         if any(failures):
-            logger.error(
-                f"{len(failures)} errors in batch (hazard_type={hazard_type.__name__}, indicator_id={indicator_id}, "
+            # only a warning: perhaps the caller does not expect data to be present for all
+            # year/scenario combinations.
+            logger.warning(
+                f"{len(failures)} requests failed in batch (hazard_type={hazard_type.__name__}, indicator_id={indicator_id}, "
                 f"scenario={scenario}, year={year}): (logs limited to first 3)"
             )
             errors = (str(i.error) for i in failures)
-            for _ in range(3):
-                logger.error(next(errors))
+            for _ in range(min(len(failures), 3)):
+                logger.warning(next(errors))
         return
 
 

diff --git a/src/physrisk/kernel/risk.py b/src/physrisk/kernel/risk.py
@@ -103,7 +103,7 @@ def _calculate_single_impact(
 class MeasureKey(NamedTuple):
     asset: Asset
     prosp_scen: str  # prospective scenario
-    year: int
+    year: Optional[int]
     hazard_type: type
 
 
@@ -238,19 +238,19 @@ def get_measure_id(
         return measure_ids_for_hazard, measure_id_lookup
 
     def calculate_risk_measures(
-        self, assets: Sequence[Asset], prosp_scens: Sequence[str], years: Sequence[int]
+        self, assets: Sequence[Asset], scenarios: Sequence[str], years: Sequence[int]
     ):
         impacts = self._calculate_all_impacts(
-            assets, prosp_scens, years, include_histo=True
+            assets, scenarios, years, include_histo=True
         )
         measures: Dict[MeasureKey, Measure] = {}
         aggregated_measures: Dict[MeasureKey, Measure] = {}
         for asset in assets:
             if type(asset) not in self._measure_calculators:
                 continue
             measure_calc = self._measure_calculators[type(asset)]
-            for prosp_scen in prosp_scens:
-                for year in years:
+            for scenario in scenarios:
+                for year in [None] if scenario == "historical" else years:
                     for hazard_type in measure_calc.supported_hazards():
                         base_impacts = impacts.get(
                             ImpactKey(
@@ -260,33 +260,35 @@ def calculate_risk_measures(
                                 key_year=None,
                             )
                         )
-                        prosp_impacts = impacts.get(
+                        # the future impact might also be the historical if that is also specified
+                        fut_impacts = impacts.get(
                             ImpactKey(
                                 asset=asset,
                                 hazard_type=hazard_type,
-                                scenario=prosp_scen,
+                                scenario=scenario,
                                 key_year=year,
                             )
                         )
+                        if base_impacts is None or fut_impacts is None:
+                            # should only happen if we are working with limited hazard scope
+                            continue
                         risk_inds = [
                             measure_calc.calc_measure(
                                 hazard_type, base_impact, prosp_impact
                             )
                             for base_impact, prosp_impact in zip(
-                                base_impacts, prosp_impacts
+                                base_impacts, fut_impacts
                             )
                         ]
                         risk_ind = [
                             risk_ind for risk_ind in risk_inds if risk_ind is not None
                         ]
                         if len(risk_ind) > 0:
                             # TODO: Aggregate  measures instead of picking the first value.
-                            measures[
-                                MeasureKey(asset, prosp_scen, year, hazard_type)
-                            ] = risk_ind[0]
-            aggregated_measures.update(
-                measure_calc.aggregate_risk_measures(
-                    measures, assets, prosp_scens, years
-                )
-            )
+                            measures[MeasureKey(asset, scenario, year, hazard_type)] = (
+                                risk_ind[0]
+                            )
+        aggregated_measures.update(
+            measure_calc.aggregate_risk_measures(measures, assets, scenarios, years)
+        )
         return impacts, aggregated_measures
diff --git a/src/physrisk/requests.py b/src/physrisk/requests.py
@@ -556,7 +556,7 @@ def _create_risk_measures(
     measures_for_assets: List[RiskMeasuresForAssets] = []
     for hazard_type in hazard_types:
         for scenario_id in scenarios:
-            for year in years:
+            for year in [None] if scenario_id == "historical" else years:
                 # we calculate and tag results for each scenario, year and hazard
                 score_key = RiskMeasureKey(
                     hazard_type=hazard_type.__name__,

diff --git a/src/physrisk/vulnerability_models/config_based_impact_curves.py b/src/physrisk/vulnerability_models/config_based_impact_curves.py
@@ -71,11 +71,11 @@ class VulnerabilityConfigItem(BaseModel):
         "curve/surface. 1 or 2 dimensional array.",
     )
     cap_of_points_x: Optional[float] = Field(
-        "", description="Cap of x (indicator/threshold)."
+        None, description="Cap of x (indicator/threshold)."
     )
-    cap_of_points_y: Optional[float] = Field("", description="Cap of y (impact).")
+    cap_of_points_y: Optional[float] = Field(None, description="Cap of y (impact).")
     activation_of_points_x: Optional[float] = Field(
-        "", description="Activation threshold of x (indicator/threshold)."
+        None, description="Activation threshold of x (indicator/threshold)."
     )
 
 

diff --git a/tests/risk_models/risk_models_test.py b/tests/risk_models/risk_models_test.py
@@ -64,7 +64,7 @@ def test_risk_indicator_model(self):
         )
         measure_ids_for_asset, definitions = model.populate_measure_definitions(assets)
         _, measures = model.calculate_risk_measures(
-            assets, prosp_scens=scenarios, years=years
+            assets, scenarios=scenarios, years=years
         )
 
         # how to get a score using the MeasureKey
@@ -293,7 +293,7 @@ def sp_precipitation(scenario, year):
         )
 
     def test_via_requests(self):
-        scenarios = ["ssp585"]
+        scenarios = ["ssp585", "historical"]
         years = [2050]
 
         assets = self._create_assets()
@@ -386,7 +386,7 @@ def test_generic_model(self):
         )
         measure_ids_for_asset, definitions = model.populate_measure_definitions(assets)
         _, measures = model.calculate_risk_measures(
-            assets, prosp_scens=scenarios, years=years
+            assets, scenarios=scenarios, years=years
         )
         np.testing.assert_approx_equal(
             measures[MeasureKey(assets[0], scenarios[0], years[0], Wind)].measure_0,