Use observation weights in score

modelskill/comparison/_collection.py

@@ -1000,6 +1000,9 @@ def score(
         """
 
         weights = kwargs.pop("weights", None)
+        if weights is None:
+            weights = {c.name: c.weight for c in self.comparers.values()}
+
         metric = _parse_metric(metric, self.metrics)
         if not (callable(metric) or isinstance(metric, str)):
             raise ValueError("metric must be a string or a function")

modelskill/comparison/_comparison.py

@@ -133,6 +133,9 @@ def _parse_dataset(data) -> xr.Dataset:
         data["Observation"].attrs["units"] = Quantity.undefined().unit
 
     data.attrs["modelskill_version"] = __version__
+
+    if "weight" not in data.attrs:
+        data.attrs["weight"] = 1.0
     return data
 
 
@@ -470,6 +473,7 @@ def from_matched_data(
         mod_items: Optional[Iterable[str | int]] = None,
         aux_items: Optional[Iterable[str | int]] = None,
         name: Optional[str] = None,
+        weight: float = 1.0,
         x: Optional[float] = None,
         y: Optional[float] = None,
         z: Optional[float] = None,
@@ -489,6 +493,7 @@ def from_matched_data(
                 z=z,
                 quantity=quantity,
             )
+            data.attrs["weight"] = weight
         return Comparer(matched_data=data, raw_mod_data=raw_mod_data)
 
     def __repr__(self):
@@ -615,11 +620,11 @@ def obs_name(self) -> str:
 
     @property
     def weight(self) -> float:
-        return self.data[self._obs_name].attrs["weight"]
+        return self.data.attrs["weight"]
 
     @weight.setter
     def weight(self, value: float) -> None:
-        self.data[self._obs_name].attrs["weight"] = value
+        self.data.attrs["weight"] = value
 
     @property
     def _unit_text(self):
@@ -833,7 +838,7 @@ def __add__(
                 matched = match_space_time(
                     observation=self._to_observation(), raw_mod_data=raw_mod_data  # type: ignore
                 )
-                cmp = self.__class__(matched_data=matched, raw_mod_data=raw_mod_data)
+                cmp = Comparer(matched_data=matched, raw_mod_data=raw_mod_data)
 
             return cmp
         else:

modelskill/matching.py

@@ -74,6 +74,7 @@ def from_matched(
     aux_items: Optional[Iterable[str | int]] = None,
     quantity: Optional[Quantity] = None,
     name: Optional[str] = None,
+    weight: float = 1.0,
     x: Optional[float] = None,
     y: Optional[float] = None,
     z: Optional[float] = None,
@@ -139,6 +140,7 @@ def from_matched(
         mod_items=mod_items,
         aux_items=aux_items,
         name=name,
+        weight=weight,
         x=x,
         y=y,
         z=z,
@@ -275,6 +277,7 @@ def _single_obs_compare(
 
     raw_mod_data = {m.name: m.extract(obs) for m in mods}
     matched_data = match_space_time(obs, raw_mod_data, max_model_gap)
+    matched_data.attrs["weight"] = obs.weight
 
     return Comparer(matched_data=matched_data, raw_mod_data=raw_mod_data)
 

modelskill/observation.py

@@ -95,23 +95,23 @@ class Observation(TimeSeries):
     def __init__(
         self,
         data: xr.Dataset,
-        weight: float = 1.0,  # TODO: cannot currently be set
+        weight: float,
         color: str = "#d62728",  # TODO: cannot currently be set
     ) -> None:
         data["time"] = self._parse_time(data.time)
 
         super().__init__(data=data)
-        self.data[self.name].attrs["weight"] = weight
-        self.data[self.name].attrs["color"] = color
+        self.data.attrs["weight"] = weight
+        self.data.attrs["color"] = color
 
     @property
     def weight(self) -> float:
         """Weighting factor for skill scores"""
-        return self.data[self.name].attrs["weight"]
+        return self.data.attrs["weight"]
 
     @weight.setter
     def weight(self, value: float) -> None:
-        self.data[self.name].attrs["weight"] = value
+        self.data.attrs["weight"] = value
 
     # TODO: move this to TimeSeries?
     @staticmethod
@@ -173,6 +173,7 @@ def __init__(
         y: Optional[float] = None,
         z: Optional[float] = None,
         name: Optional[str] = None,
+        weight: float = 1.0,
         quantity: Optional[Quantity] = None,
         aux_items: Optional[list[int | str]] = None,
         attrs: Optional[dict] = None,
@@ -194,7 +195,7 @@ def __init__(
         _validate_attrs(data.attrs, attrs)
         data.attrs = {**data.attrs, **(attrs or {})}
 
-        super().__init__(data=data)
+        super().__init__(data=data, weight=weight)
 
     @property
     def geometry(self):
@@ -316,6 +317,7 @@ def __init__(
         *,
         item: Optional[int | str] = None,
         name: Optional[str] = None,
+        weight: float = 1.0,
         x_item: Optional[int | str] = 0,
         y_item: Optional[int | str] = 1,
         keep_duplicates: bool | str = "first",
@@ -350,7 +352,7 @@ def __init__(
         _validate_attrs(data.attrs, attrs)
         data.attrs = {**data.attrs, **(attrs or {})}
 
-        super().__init__(data=data)
+        super().__init__(data=data, weight=weight)
 
     def __repr__(self):
         out = f"TrackObservation: {self.name}, n={self.n_points}"

modelskill/plotting/_misc.py

@@ -3,6 +3,7 @@
 from typing import Optional, Sequence, Tuple, Union
 
 import matplotlib.pyplot as plt
+from matplotlib.axes import Axes
 import numpy as np
 import pandas as pd
 
@@ -16,7 +17,7 @@ def _get_ax(ax=None, figsize=None):
     return ax
 
 
-def _get_fig_ax(ax: plt.Axes | None = None, figsize=None):
+def _get_fig_ax(ax: Axes | None = None, figsize=None):
     if ax is None:
         fig, ax = plt.subplots(figsize=figsize)
     else:

modelskill/plotting/_scatter.py

@@ -549,7 +549,7 @@ def _plot_summary_table(
     for ti in text_:
         text_col_i = fig.text(x + dx, 0.6, ti, **txt_settings)
         ## Render, and get width
-        plt.draw()
+        # plt.draw() # TOOO this causes an error and I have no idea why it is here
         dx = (
             dx
             + figure_transform.inverted().transform(

tests/test_comparercollection.py

@@ -199,6 +199,11 @@ def test_save_and_load_preserves_raw_model_data(cc, tmp_path):
     assert len(cc2["fake point obs"].raw_mod_data["m1"]) == 6
 
 
+def test_scatter(cc):
+    ax = cc.plot.scatter(skill_table=True)
+    assert ax is not None
+
+
 def test_hist(cc):
     ax = cc.sel(model="m1").plot.hist()
     assert ax is not None

tests/test_pointcompare.py

@@ -112,27 +112,85 @@ def test_score(modelresult_oresund_WL, klagshamn, drogden):
     s.root_mean_squared_error.data.mean() == pytest.approx(0.1986296276629835)
 
 
-# def test_weighted_score(modelresult_oresund_WL, klagshamn, drogden):
-#     mr = modelresult_oresund_WL
+def test_weighted_score(modelresult_oresund_WL):
+    o1 = ms.PointObservation(
+        "tests/testdata/smhi_2095_klagshamn.dfs0",
+        item=0,
+        x=366844,
+        y=6154291,
+        name="Klagshamn",
+    )
+    o2 = ms.PointObservation(
+        "tests/testdata/dmi_30357_Drogden_Fyr.dfs0",
+        item=0,
+        x=355568.0,
+        y=6156863.0,
+        quantity=ms.Quantity(
+            "Water Level", unit="meter"
+        ),  # not sure if this is relevant in this test
+    )
+
+    mr = ms.model_result("tests/testdata/Oresund2D_subset.dfsu", item=0, name="Oresund")
+
+    cc = ms.match(obs=[o1, o2], mod=mr)
+    unweighted = cc.score()
+    assert unweighted["Oresund"] == pytest.approx(0.1986296276629835)
+
+    # Weighted
+
+    o1_w = ms.PointObservation(
+        "tests/testdata/smhi_2095_klagshamn.dfs0",
+        item=0,
+        x=366844,
+        y=6154291,
+        name="Klagshamn",
+        weight=10.0,
+    )
+
+    o2_w = ms.PointObservation(
+        "tests/testdata/dmi_30357_Drogden_Fyr.dfs0",
+        item=0,
+        x=355568.0,
+        y=6156863.0,
+        quantity=ms.Quantity(
+            "Water Level", unit="meter"
+        ),  # not sure if this is relevant in this test
+        weight=0.1,
+    )
 
-#     cc = ms.match([klagshamn, drogden], mr)
-#     unweighted_skill = cc.score()
+    cc_w = ms.match(obs=[o1_w, o2_w], mod=mr)
+    weighted = cc_w.score()
 
-#     con = ms.Connector()
+    assert weighted["Oresund"] == pytest.approx(0.1666888485806514)
 
-#     con.add(klagshamn, mr, weight=0.9, validate=False)
-#     con.add(drogden, mr, weight=0.1, validate=False)
-#     cc = con.extract()
-#     weighted_skill = cc.score()
-#     assert unweighted_skill != weighted_skill
 
-#     obs = [klagshamn, drogden]
+def test_weighted_score_from_prematched():
+    df = pd.DataFrame(
+        {"Oresund": [0.0, 1.0], "klagshamn": [0.0, 1.0], "drogden": [-1.0, 2.0]}
+    )
+
+    cmp1 = ms.from_matched(
+        df[["Oresund", "klagshamn"]],
+        mod_items=["Oresund"],
+        obs_item="klagshamn",
+        weight=100.0,
+    )
+    cmp2 = ms.from_matched(
+        df[["Oresund", "drogden"]],
+        mod_items=["Oresund"],
+        obs_item="drogden",
+        weight=0.0,
+    )
+    assert cmp1.weight == 100.0
+    assert cmp2.weight == 0.0
+    assert cmp1.score()["Oresund"] == pytest.approx(0.0)
+    assert cmp2.score()["Oresund"] == pytest.approx(1.0)
 
-#     con = ms.Connector(obs, mr, weight=[0.9, 0.1], validate=False)
-#     cc = con.extract()
-#     weighted_skill2 = cc.score()
+    cc = ms.ComparerCollection([cmp1, cmp2])
+    assert cc["klagshamn"].weight == 100.0
+    assert cc["drogden"].weight == 0.0
 
-#     assert weighted_skill == weighted_skill2
+    assert cc.score()["Oresund"] == pytest.approx(1.0)
 
 
 def test_misc_properties(klagshamn, drogden):