Merge pull request #269 from DHI/matching-refactor

More work on internal data structures
DHI · Nov 19, 2023 · d09b54b · d09b54b
2 parents cbf562d + 5bf3ea3
commit d09b54b
Show file tree

Hide file tree

Showing 46 changed files with 6,079 additions and 2,353 deletions.
diff --git a/docs/simple-compare.md b/docs/simple-compare.md
@@ -47,7 +47,7 @@ c = ms.compare(fn_obs, fn_mod, mod_item=0)
 ### 4. Analysis and plotting
 
 The returned
-[PointComparer](modelskill.comparison.PointComparer) can make
+[Comparer](modelskill.comparison.Comparer) can make
 scatter plots, skill assessment, time series plots etc.
 
 ```python

diff --git a/modelskill/__init__.py b/modelskill/__init__.py
@@ -29,7 +29,7 @@
 from .model import PointModelResult, TrackModelResult, GridModelResult, DfsuModelResult
 from .observation import PointObservation, TrackObservation
 from .matching import compare, from_matched
-from .connection import Connector
+from .connection import Connector, from_config
 from .settings import options, get_option, set_option, reset_option, load_style
 from . import plotting
 from .comparison import ComparerCollection
@@ -77,11 +77,3 @@ def load(filename: Union[str, Path]) -> ComparerCollection:
     "plotting",
     "from_config",
 ]
-
-
-def from_config(
-    configuration: Union[dict, str], *, validate_eum=True, relative_path=True
-):
-    return Connector.from_config(
-        configuration, validate_eum=validate_eum, relative_path=relative_path
-    )
diff --git a/modelskill/comparison/__init__.py b/modelskill/comparison/__init__.py
@@ -1,7 +1,7 @@
-"""The `comparison` module contains different types of comparer classes for
-fixed locations (PointComparer), or locations moving in space (TrackComparer).
+"""The `comparison` module contains different types of classes for single 
+observation comparison (Comparer), and collections of Comparers (ComparerCollection).
 """
-from ._comparison import Comparer, PointComparer, TrackComparer
+from ._comparison import Comparer
 from ._collection import ComparerCollection
 
-__all__ = ["Comparer", "PointComparer", "TrackComparer", "ComparerCollection"]
+__all__ = ["Comparer", "ComparerCollection"]
diff --git a/modelskill/comparison/_collection.py b/modelskill/comparison/_collection.py
@@ -17,14 +17,14 @@
 from ..settings import options, reset_option
 
 from ..utils import _get_idx, _get_name
-from ._comparison import (
-    Comparer,
-    IdOrNameTypes,
-    TimeTypes,
+from ._comparison import Comparer
+from ._utils import (
     _parse_metric,
-    _parse_groupby,
-    _groupby_df,
     _add_spatial_grid_to_df,
+    _groupby_df,
+    _parse_groupby,
+    IdOrNameTypes,
+    TimeTypes,
 )
 from ._comparison import _get_deprecated_args  # TODO remove in v 1.1
 
@@ -97,7 +97,7 @@ def add_comparer(self, comparer: Union["Comparer", "ComparerCollection"]) -> Non
 
         Parameters
         ----------
-        comparer : (PointComparer, TrackComparer, ComparerCollection)
+        comparer : (Comparer, ComparerCollection)
             Comparer to add to this collection
         """
         if isinstance(comparer, (ComparerCollection, Sequence)):
@@ -181,7 +181,7 @@ def var_names(self) -> List[str]:
         """List of unique variable names"""
         unique_names = []
         for cmp in self.comparers.values():
-            n = cmp.quantity_name
+            n = cmp.quantity.name
             if n not in unique_names:
                 unique_names.append(n)
         return unique_names
@@ -211,12 +211,13 @@ def to_dataframe(self) -> pd.DataFrame:
         for cmp in self.comparers.values():
             for j in range(cmp.n_models):
                 mod_name = cmp.mod_names[j]
-                df = cmp.data[[mod_name]].to_dataframe().copy()
-                df.columns = ["mod_val"]
+                # drop "x", "y",  ?
+                df = cmp.data.drop_vars(["z"])[[mod_name]].to_dataframe().copy()
+                df = df.rename(columns={mod_name: "mod_val"})
                 df["model"] = mod_name
                 df["observation"] = cmp.name
                 if self.n_variables > 1:
-                    df["variable"] = cmp.quantity_name
+                    df["variable"] = cmp.quantity.name
                 df["x"] = cmp.x
                 df["y"] = cmp.y
                 df["obs_val"] = cmp.obs
@@ -329,7 +330,7 @@ def sel(
 
         cc = ComparerCollection()
         for cmp in self.comparers.values():
-            if cmp.name in observation and cmp.quantity_name in variable:
+            if cmp.name in observation and cmp.quantity.name in variable:
                 thismodel = (
                     [m for m in mod_names if m in cmp.mod_names] if model else None
                 )