fix,doc(interpolation): minor improvements and fixes

Author: spjuhel

climada/trajectories/interpolation.py

@@ -50,6 +50,9 @@ def exponential_interp_imp_mat(mat_start, mat_end, interpolation_range, rate) ->
     `mat_start` and `mat_end`.
     """
     # Convert matrices to logarithmic domain
+    if rate <= 0:
+        raise ValueError("Rate for exponential interpolation must be positive")
+
     mat_start = mat_start.copy()
     mat_end = mat_end.copy()
     mat_start.data = np.log(mat_start.data + np.finfo(float).eps) / np.log(rate)
@@ -65,8 +68,16 @@ def exponential_interp_imp_mat(mat_start, mat_end, interpolation_range, rate) ->
     return res
 
 
-def linear_interp_arrays(arr_start, arr_end, interpolation_range):
-    """Perform linear interpolation between two arrays (of a scalar metric) over an interpolation range."""
+def linear_interp_arrays(arr_start, arr_end):
+    """Perform linear interpolation between two arrays of `n` dates of one or multiple scalar metrics.
+
+    Returns a `n` sized arrays where the values linearly change from `arr_start` to `arr_end` over the `n` dates.
+    """
+    if arr_start.shape != arr_end.shape:
+        raise ValueError(
+            f"Cannot interpolate arrays of different shapes: {arr_start.shape} and {arr_end.shape}."
+        )
+    interpolation_range = arr_start.shape[0]
     prop1 = np.linspace(0, 1, interpolation_range)
     prop0 = 1 - prop1
     if arr_start.ndim > 1:
@@ -75,32 +86,35 @@ def linear_interp_arrays(arr_start, arr_end, interpolation_range):
     return np.multiply(arr_start, prop0) + np.multiply(arr_end, prop1)
 
 
-def exponential_interp_arrays(arr_start, arr_end, interpolation_range, rate):
-    """Perform exponential interpolation between two arrays (of a scalar metric) over an interpolation range with a growth rate `rate`."""
+def exponential_interp_arrays(arr_start, arr_end, rate):
+    """Perform exponential interpolation between two arrays of `n` dates of one or multiple scalar metrics.
+
+    Returns a `n` sized arrays where the values exponentially change from `arr_start` to `arr_end` over the `n` dates.
+    """
+
+    if rate <= 0:
+        raise ValueError("Rate for exponential interpolation must be positive")
+
+    if arr_start.shape != arr_end.shape:
+        raise ValueError(
+            f"Cannot interpolate arrays of different shapes: {arr_start.shape} and {arr_end.shape}."
+        )
+    interpolation_range = arr_start.shape[0]
+
     prop1 = np.linspace(0, 1, interpolation_range)
     prop0 = 1 - prop1
     if arr_start.ndim > 1:
         prop0, prop1 = prop0.reshape(-1, 1), prop1.reshape(-1, 1)
 
     return np.exp(
         (
-            np.multiply(np.log(arr_start) / np.log(rate), prop0)
-            + np.multiply(np.log(arr_end) / np.log(rate), prop1)
+            np.multiply(np.log(arr_start + np.finfo(float).eps) / np.log(rate), prop0)
+            + np.multiply(np.log(arr_end + np.finfo(float).eps) / np.log(rate), prop1)
         )
         * np.log(rate)
     )
 
 
-def logarithmic_interp_arrays(arr_start, arr_end, interpolation_range):
-    """Perform logarithmic (natural logarithm) interpolation between two arrays (of a scalar metric) over an interpolation range."""
-    prop1 = np.logspace(0, 1, interpolation_range)
-    prop0 = 1 - prop1
-    if arr_start.ndim > 1:
-        prop0, prop1 = prop0.reshape(-1, 1), prop1.reshape(-1, 1)
-
-    return np.multiply(arr_start, prop0) + np.multiply(arr_end, prop1)
-
-
 class InterpolationStrategyBase(ABC):
     exposure_interp: Callable
     hazard_interp: Callable
@@ -126,17 +140,11 @@ def interp_exposure_dim(
 
         return res
 
-    def interp_hazard_dim(
-        self, metric_0, metric_1, interpolation_range: int, **kwargs
-    ) -> np.ndarray:
-        return self.hazard_interp(metric_0, metric_1, interpolation_range, **kwargs)
+    def interp_hazard_dim(self, metric_0, metric_1, **kwargs) -> np.ndarray:
+        return self.hazard_interp(metric_0, metric_1, **kwargs)
 
-    def interp_vulnerability_dim(
-        self, metric_0, metric_1, interpolation_range: int, **kwargs
-    ) -> np.ndarray:
-        return self.vulnerability_interp(
-            metric_0, metric_1, interpolation_range, **kwargs
-        )
+    def interp_vulnerability_dim(self, metric_0, metric_1, **kwargs) -> np.ndarray:
+        return self.vulnerability_interp(metric_0, metric_1, **kwargs)
 
 
 class InterpolationStrategy(InterpolationStrategyBase):

climada/trajectories/riskperiod.py

@@ -671,13 +671,13 @@ def calc_eai(self):
             self.per_date_eai_H1V1,
         )
         per_date_eai_V0 = self.interpolation_strategy.interp_hazard_dim(
-            per_date_eai_H0V0, per_date_eai_H1V0, self.time_points
+            per_date_eai_H0V0, per_date_eai_H1V0
         )
         per_date_eai_V1 = self.interpolation_strategy.interp_hazard_dim(
-            per_date_eai_H0V1, per_date_eai_H1V1, self.time_points
+            per_date_eai_H0V1, per_date_eai_H1V1
         )
         per_date_eai = self.interpolation_strategy.interp_vulnerability_dim(
-            per_date_eai_V0, per_date_eai_V1, self.time_points
+            per_date_eai_V0, per_date_eai_V1
         )
         return per_date_eai
 
@@ -690,6 +690,7 @@ def calc_eai_gdf(self):
         eai_gdf["coord_id"] = eai_gdf.index
         eai_gdf = eai_gdf.merge(df, on="coord_id")
         eai_gdf = eai_gdf.rename(columns={"group_id": "group"})
+        eai_gdf["group"] = eai_gdf["group"].astype("category")
         eai_gdf["metric"] = "eai"
         eai_gdf["measure"] = self.measure.name if self.measure else "no_measure"
         return eai_gdf
@@ -720,14 +721,17 @@ def calc_aai_per_group_metric(self):
                 "Group id are changing between present and future snapshot. Per group AAI will be linearly interpolated."
             )
             eai_fut_groups = self.eai_gdf.copy()
-            eai_fut_groups.index = self._group_id_E1
-            aai_fut_groups = (
-                eai_fut_groups.groupby(["date", "group"], as_index=False)["risk"]
-                .sum()
-                .values()
+            eai_fut_groups["group"] = pd.Categorical(
+                np.tile(self._group_id_E1, len(self.date_idx)),
+                categories=np.unique(
+                    np.concatenate([self._group_id_E0, self._group_id_E1])
+                ),
             )
+            aai_fut_groups = eai_fut_groups.groupby(["date", "group"], as_index=False)[
+                "risk"
+            ].sum()
             aai_per_group_df["risk"] = linear_interp_arrays(
-                aai_per_group_df["risk"], aai_fut_groups, self.time_points
+                aai_per_group_df["risk"].values, aai_fut_groups["risk"].values
             )
 
         aai_per_group_df["metric"] = "aai"
@@ -745,13 +749,13 @@ def calc_return_periods_metric(self, return_periods):
             self.per_date_return_periods_H1V1(return_periods),
         )
         per_date_rp_V0 = self.interpolation_strategy.interp_hazard_dim(
-            per_date_rp_H0V0, per_date_rp_H1V0, self.time_points
+            per_date_rp_H0V0, per_date_rp_H1V0
         )
         per_date_rp_V1 = self.interpolation_strategy.interp_hazard_dim(
-            per_date_rp_H0V1, per_date_rp_H1V1, self.time_points
+            per_date_rp_H0V1, per_date_rp_H1V1
         )
         per_date_rp = self.interpolation_strategy.interp_vulnerability_dim(
-            per_date_rp_V0, per_date_rp_V1, self.time_points
+            per_date_rp_V0, per_date_rp_V1
         )
         rp_df = pd.DataFrame(
             index=self.date_idx, columns=return_periods, data=per_date_rp
@@ -763,11 +767,11 @@ def calc_return_periods_metric(self, return_periods):
         return rp_df
 
     def calc_risk_components_metric(self):
-        per_date_aai_V0 = self.interpolation_strategy.interp_vulnerability_dim(
-            self.per_date_aai_H0V0, self.per_date_aai_H1V0, self.time_points
+        per_date_aai_V0 = self.interpolation_strategy.interp_hazard_dim(
+            self.per_date_aai_H0V0, self.per_date_aai_H1V0
         )
         per_date_aai_H0 = self.interpolation_strategy.interp_vulnerability_dim(
-            self.per_date_aai_H0V0, self.per_date_aai_H0V1, self.time_points
+            self.per_date_aai_H0V0, self.per_date_aai_H0V1
         )
         df = pd.DataFrame(
             {