Fixed types

Author: lucasimi

src/tdamapper/_common.py

@@ -27,7 +27,7 @@ def deprecated(msg: str) -> Callable[..., Any]:
     """
 
     def deprecated_func(func: Callable[..., Any]) -> Callable[..., Any]:
-        def wrapper(*args: list[Any], **kwargs: dict[str, Any]) -> Any:
+        def wrapper(*args: list[Any], **kwargs: Any) -> Any:
             warnings.warn(msg, DeprecationWarning, stacklevel=2)
             return func(*args, **kwargs)
 
@@ -213,7 +213,7 @@ def profile(n_lines: int = 10) -> Callable[..., Any]:
     """
 
     def decorator(func: Callable[..., Any]) -> Callable[..., Any]:
-        def wrapper(*args: list[Any], **kwargs: dict[str, Any]) -> Any:
+        def wrapper(*args: list[Any], **kwargs: Any) -> Any:
             profiler = cProfile.Profile()
             profiler.enable()
             result = func(*args, **kwargs)

src/tdamapper/app.py

@@ -202,7 +202,7 @@ def _umap(X: NDArray[np.float_]) -> NDArray[np.float_]:
 
 
 def run_mapper(
-    df: pd.DataFrame, **kwargs: dict[str, Any]
+    df: pd.DataFrame, **kwargs: Any
 ) -> Optional[tuple[nx.Graph, pd.DataFrame]]:
     """
     Runs the Mapper algorithm on the provided DataFrame and returns the Mapper
@@ -301,7 +301,7 @@ def create_mapper_figure(
     df_y: pd.DataFrame,
     df_target: pd.DataFrame,
     mapper_graph: nx.Graph,
-    **kwargs: dict[str, Any],
+    **kwargs: Any,
 ) -> go.Figure:
     """
     Renders the Mapper graph as a Plotly figure.

src/tdamapper/utils/metrics.py

@@ -58,7 +58,7 @@ def get_supported_metrics() -> list[MetricLiteral]:
     return list(get_args(MetricLiteral))
 
 
-def euclidean(**kwargs: dict[str, Any]) -> Metric[Any]:
+def euclidean() -> Metric[Any]:
     """
     Return the Euclidean distance function for vectors.
 
@@ -70,7 +70,7 @@ def euclidean(**kwargs: dict[str, Any]) -> Metric[Any]:
     return _metrics.euclidean
 
 
-def manhattan(**kwargs: dict[str, Any]) -> Metric[Any]:
+def manhattan() -> Metric[Any]:
     """
     Return the Manhattan distance function for vectors.
 
@@ -82,7 +82,7 @@ def manhattan(**kwargs: dict[str, Any]) -> Metric[Any]:
     return _metrics.manhattan
 
 
-def chebyshev(**kwargs: dict[str, Any]) -> Metric[Any]:
+def chebyshev() -> Metric[Any]:
     """
     Return the Chebyshev distance function for vectors.
 
@@ -94,7 +94,7 @@ def chebyshev(**kwargs: dict[str, Any]) -> Metric[Any]:
     return _metrics.chebyshev
 
 
-def minkowski(**kwargs: dict[str, Any]) -> Metric[Any]:
+def minkowski(p: Union[int, float]) -> Metric[Any]:
     """
     Return the Minkowski distance function for order p on vectors.
 
@@ -106,9 +106,6 @@ def minkowski(**kwargs: dict[str, Any]) -> Metric[Any]:
     :param p: The order of the Minkowski distance.
     :return: The Minkowski distance function.
     """
-    p = kwargs.get("p", 2)
-    if not isinstance(p, (int, float)):
-        raise TypeError("p must be an integer or a float")
     if p == 1:
         return manhattan()
     if p == 2:
@@ -122,7 +119,7 @@ def dist(x: Any, y: Any) -> float:
     return dist
 
 
-def cosine(**kwargs: dict[str, Any]) -> Metric[Any]:
+def cosine() -> Metric[Any]:
     """
     Return the cosine distance function for vectors.
 
@@ -145,9 +142,7 @@ def cosine(**kwargs: dict[str, Any]) -> Metric[Any]:
     return _metrics.cosine
 
 
-def get_metric(
-    metric: Union[MetricLiteral, Metric[Any]], **kwargs: dict[str, Any]
-) -> Metric[Any]:
+def get_metric(metric: Union[MetricLiteral, Metric[Any]], **kwargs: Any) -> Metric[Any]:
     """
     Return a distance function based on the specified string or callable.
 

tests/test_unit_metrics.py

@@ -31,6 +31,34 @@
 RANDOM = dataset_random(2, 100)
 
 
+def _check_values(m1, m2, a, b):
+    m1_div_by_zero = False
+    m1_is_nan = False
+
+    m2_div_by_zero = False
+    m2_is_nan = False
+
+    try:
+        m1_value = m1(a, b)
+        if np.isnan(m1_value):
+            m1_is_nan = True
+    except ZeroDivisionError:
+        m1_div_by_zero = True
+    try:
+        m2_value = m2(a, b)
+        if np.isnan(m2_value):
+            m2_is_nan = True
+    except ZeroDivisionError:
+        m2_div_by_zero = True
+    assert m1_div_by_zero == m2_div_by_zero
+    assert m1_is_nan == m2_is_nan
+    if m1_div_by_zero or m2_div_by_zero:
+        return True
+    if m1_is_nan or m2_is_nan:
+        return True
+    return math.isclose(m1_value, m2_value)
+
+
 @pytest.mark.parametrize("data", [SIMPLE, TWO_LINES, GRID, RANDOM])
 @pytest.mark.parametrize(
     "m1, m2",
@@ -39,28 +67,14 @@
         (manhattan(), get_metric("manhattan")),
         (chebyshev(), get_metric("chebyshev")),
         (minkowski(p=3), get_metric("minkowski", p=3)),
+        (minkowski(p=2.5), get_metric("minkowski", p=2.5)),
         (cosine(), get_metric("cosine")),
     ],
 )
 def test_metrics(m1, m2, data):
     for a in data:
         for b in data:
-            m1_fail = False
-            m2_fail = False
-            m1_value = 0.0
-            m2_value = 0.0
-            try:
-                m1_value = m1(a, b)
-            except Exception:
-                m1_fail = True
-            try:
-                m2_value = m2(a, b)
-            except Exception:
-                m2_fail = True
-            assert m1_fail == m2_fail
-            if np.isnan(m1_value) and np.isnan(m2_value):
-                return
-            assert math.isclose(m1_value, m2_value)
+            assert _check_values(m1, m2, a, b)
 
 
 def test_supported_metrics():