Flattened hierarchies. Added tests

lucasimi · lucasimi · commit 0c9e85303bb8 · 2025-06-28T07:18:09.000+02:00
diff --git a/docs/source/apiref.rst b/docs/source/apiref.rst
@@ -33,18 +33,18 @@ tdamapper.clustering
    :undoc-members:
    :show-inheritance:
 
-tdamapper.utils.metrics
+tdamapper.metrics
 -----------------------
 
-.. automodule:: tdamapper.utils.metrics
+.. automodule:: tdamapper.metrics
    :members:
    :undoc-members:
    :show-inheritance:
 
-tdamapper.utils.vptree
+tdamapper.vptree
 -----------------------
 
-.. automodule:: tdamapper.utils.vptree
+.. automodule:: tdamapper.vptree
    :members:
    :undoc-members:
    :show-inheritance:
diff --git a/src/tdamapper/_metrics.py b/src/tdamapper/_metrics.py
diff --git a/src/tdamapper/core.py b/src/tdamapper/core.py
@@ -39,7 +39,7 @@
 from numpy.typing import NDArray
 
 from tdamapper._common import EstimatorMixin, ParamsMixin, clone, deprecated
-from tdamapper.utils.unionfind import UnionFind
+from tdamapper.unionfind import UnionFind
 
 ATTR_IDS = "ids"
 
diff --git a/src/tdamapper/cover.py b/src/tdamapper/cover.py
@@ -90,11 +90,11 @@ class BallCover(ParamsMixin):
         Defaults to 1.0.
     :type radius: float
     :param metric: The metric used to define the distance between points.
-        Accepts any value compatible with `tdamapper.utils.metrics.get_metric`.
+        Accepts any value compatible with `tdamapper.metrics.get_metric`.
         Defaults to 'euclidean'.
     :type metric: str or callable
     :param metric_params: Additional parameters for the metric function, to be
-        passed to `tdamapper.utils.metrics.get_metric`. Defaults to None.
+        passed to `tdamapper.metrics.get_metric`. Defaults to None.
     :type metric_params: dict, optional
     :param kind: Specifies whether to use a flat or a hierarchical vantage
         point tree. Acceptable values are 'flat' or 'hierarchical'. Defaults
@@ -187,11 +187,11 @@ class KNNCover(ParamsMixin):
         Defaults to 1.
     :type neighbors: int
     :param metric: The metric used to define the distance between points.
-        Accepts any value compatible with `tdamapper.utils.metrics.get_metric`.
+        Accepts any value compatible with `tdamapper.metrics.get_metric`.
         Defaults to 'euclidean'.
     :type metric: str or callable
     :param metric_params: Additional parameters for the metric function, to be
-        passed to `tdamapper.utils.metrics.get_metric`. Defaults to None.
+        passed to `tdamapper.metrics.get_metric`. Defaults to None.
     :type metric_params: dict, optional
     :param kind: Specifies whether to use a flat or a hierarchical vantage
         point tree. Acceptable values are 'flat' or 'hierarchical'. Defaults
diff --git a/src/tdamapper/heap.py b/src/tdamapper/heap.py
diff --git a/src/tdamapper/metrics.py b/src/tdamapper/metrics.py
@@ -30,7 +30,7 @@
 
 import numpy as np
 
-import tdamapper.utils._metrics as _metrics
+import tdamapper._metrics as _metrics
 
 _EUCLIDEAN = "euclidean"
 _MANHATTAN = "manhattan"
diff --git a/src/tdamapper/plot.py b/src/tdamapper/plot.py
@@ -7,9 +7,9 @@
 import numpy as np
 
 from tdamapper._common import deprecated
-from tdamapper.plot_backends.plot_matplotlib import plot_matplotlib
-from tdamapper.plot_backends.plot_plotly import plot_plotly, plot_plotly_update
-from tdamapper.plot_backends.plot_pyvis import plot_pyvis
+from tdamapper.plot_matplotlib import plot_matplotlib
+from tdamapper.plot_plotly import plot_plotly, plot_plotly_update
+from tdamapper.plot_pyvis import plot_pyvis
 
 
 class MapperPlot:
diff --git a/src/tdamapper/plot_matplotlib.py b/src/tdamapper/plot_matplotlib.py
diff --git a/src/tdamapper/plot_plotly.py b/src/tdamapper/plot_plotly.py
diff --git a/src/tdamapper/plot_pyvis.py b/src/tdamapper/plot_pyvis.py
diff --git a/src/tdamapper/quickselect.py b/src/tdamapper/quickselect.py
diff --git a/src/tdamapper/search.py b/src/tdamapper/search.py
@@ -9,14 +9,14 @@
 from __future__ import annotations
 
 import math
-from typing import Any, Callable, Dict, Generator, List, Optional, Union
+from typing import Any, Callable, Dict, List, Optional, Union
 
 import numpy as np
 
 from tdamapper._common import warn_user
 from tdamapper.core import ArrayLike, PointLike
-from tdamapper.utils.metrics import chebyshev, get_metric
-from tdamapper.utils.vptree import VPTree
+from tdamapper.metrics import chebyshev, get_metric
+from tdamapper.vptree import VPTree
 
 
 class _Pullback:
@@ -346,15 +346,15 @@ def fit(self, X: ArrayLike) -> CubicalSearch:
             warn_user("The parameter overlap_frac is expected to be <= 0.5")
         self._min, self._max, self._delta = self._get_bounds(X)
         radius = 1.0 / (2.0 - 2.0 * self._overlap_frac)
-        self._cover = BallSearch(
+        self._ball_search = BallSearch(
             radius,
             metric=_Pullback(self._gamma_n, chebyshev()),
             kind=self.kind,
             leaf_capacity=self.leaf_capacity,
             leaf_radius=self.leaf_radius,
             pivoting=self.pivoting,
         )
-        self._cover.fit(X)
+        self._ball_search.fit(X)
         return self
 
     def search(self, x: PointLike) -> List[int]:
@@ -370,7 +370,7 @@ def search(self, x: PointLike) -> List[int]:
         :rtype: list[int]
         """
         center = self._phi(x)
-        return self._cover.search(center)
+        return self._ball_search.search(center)
 
 
 class CubicalLandmarks(CubicalSearch):
diff --git a/src/tdamapper/unionfind.py b/src/tdamapper/unionfind.py
diff --git a/src/tdamapper/utils/vptree_flat/__init__.py b/src/tdamapper/utils/vptree_flat/__init__.py
diff --git a/src/tdamapper/utils/vptree_hier/__init__.py b/src/tdamapper/utils/vptree_hier/__init__.py
diff --git a/src/tdamapper/vptree.py b/src/tdamapper/vptree.py
@@ -2,8 +2,8 @@
 A module for fast knn and range searches, depending only on a given metric
 """
 
-from tdamapper.utils.vptree_flat.vptree import VPTree as FVPT
-from tdamapper.utils.vptree_hier.vptree import VPTree as HVPT
+from tdamapper.vptree_flat.vptree import VPTree as FVPT
+from tdamapper.vptree_hier.vptree import VPTree as HVPT
 
 
 class VPTree:
@@ -13,11 +13,11 @@ class VPTree:
     :param X: A dataset of n points.
     :type X: array-like of shape (n, m) or list-like of length n
     :param metric: The metric used to define the distance between points.
-        Accepts any value compatible with `tdamapper.utils.metrics.get_metric`.
+        Accepts any value compatible with `tdamapper.metrics.get_metric`.
         Defaults to 'euclidean'.
     :type metric: str or callable
     :param metric_params: Additional parameters for the metric function, to be
-        passed to `tdamapper.utils.metrics.get_metric`. Defaults to None.
+        passed to `tdamapper.metrics.get_metric`. Defaults to None.
     :type metric_params: dict, optional
     :param kind: Specifies whether to use a flat or a hierarchical vantage
         point tree. Acceptable values are 'flat' or 'hierarchical'. Defaults
diff --git a/src/tdamapper/vptree_flat/__init__.py b/src/tdamapper/vptree_flat/__init__.py
diff --git a/src/tdamapper/vptree_flat/ball_search.py b/src/tdamapper/vptree_flat/ball_search.py
@@ -1,4 +1,4 @@
-from tdamapper.utils.vptree_flat.common import _mid
+from tdamapper.vptree_flat.common import _mid
 
 
 class BallSearch:
diff --git a/src/tdamapper/vptree_flat/builder.py b/src/tdamapper/vptree_flat/builder.py
@@ -2,7 +2,7 @@
 
 import numpy as np
 
-from tdamapper.utils.vptree_flat.common import VPArray
+from tdamapper.vptree_flat.common import VPArray
 
 
 def _mid(start, end):
diff --git a/src/tdamapper/vptree_flat/common.py b/src/tdamapper/vptree_flat/common.py
@@ -1,4 +1,4 @@
-from tdamapper.utils.quickselect import quickselect, swap_all
+from tdamapper.quickselect import quickselect, swap_all
 
 
 def _mid(start, end):
diff --git a/src/tdamapper/vptree_flat/knn_search.py b/src/tdamapper/vptree_flat/knn_search.py
@@ -1,5 +1,5 @@
-from tdamapper.utils.heap import MaxHeap
-from tdamapper.utils.vptree_flat.common import _mid
+from tdamapper.heap import MaxHeap
+from tdamapper.vptree_flat.common import _mid
 
 _PRE = 0
 _POST = 1
diff --git a/src/tdamapper/vptree_flat/vptree.py b/src/tdamapper/vptree_flat/vptree.py
@@ -1,7 +1,7 @@
-from tdamapper.utils.metrics import get_metric
-from tdamapper.utils.vptree_flat.ball_search import BallSearch
-from tdamapper.utils.vptree_flat.builder import Builder
-from tdamapper.utils.vptree_flat.knn_search import KnnSearch
+from tdamapper.metrics import get_metric
+from tdamapper.vptree_flat.ball_search import BallSearch
+from tdamapper.vptree_flat.builder import Builder
+from tdamapper.vptree_flat.knn_search import KnnSearch
 
 
 class VPTree:
diff --git a/src/tdamapper/vptree_hier/__init__.py b/src/tdamapper/vptree_hier/__init__.py
diff --git a/src/tdamapper/vptree_hier/ball_search.py b/src/tdamapper/vptree_hier/ball_search.py
diff --git a/src/tdamapper/vptree_hier/builder.py b/src/tdamapper/vptree_hier/builder.py
@@ -2,7 +2,7 @@
 
 import numpy as np
 
-from tdamapper.utils.vptree_hier.common import Leaf, Node, VPArray, _mid
+from tdamapper.vptree_hier.common import Leaf, Node, VPArray, _mid
 
 
 class Builder:
diff --git a/src/tdamapper/vptree_hier/common.py b/src/tdamapper/vptree_hier/common.py
@@ -1,4 +1,4 @@
-from tdamapper.utils.quickselect import quickselect, swap_all
+from tdamapper.quickselect import quickselect, swap_all
 
 
 def _mid(start, end):
diff --git a/src/tdamapper/vptree_hier/knn_search.py b/src/tdamapper/vptree_hier/knn_search.py
@@ -1,4 +1,4 @@
-from tdamapper.utils.heap import MaxHeap
+from tdamapper.heap import MaxHeap
 
 
 class KnnSearch:
diff --git a/src/tdamapper/vptree_hier/vptree.py b/src/tdamapper/vptree_hier/vptree.py
@@ -1,7 +1,7 @@
-from tdamapper.utils.metrics import get_metric
-from tdamapper.utils.vptree_hier.ball_search import BallSearch
-from tdamapper.utils.vptree_hier.builder import Builder
-from tdamapper.utils.vptree_hier.knn_search import KnnSearch
+from tdamapper.metrics import get_metric
+from tdamapper.vptree_hier.ball_search import BallSearch
+from tdamapper.vptree_hier.builder import Builder
+from tdamapper.vptree_hier.knn_search import KnnSearch
 
 
 class VPTree:
diff --git a/tests/test_bench_cover.py b/tests/test_bench_cover.py
@@ -5,9 +5,9 @@
 from sklearn.datasets import load_digits
 
 from tdamapper._common import profile
-from tdamapper.utils.metrics import euclidean
-from tdamapper.utils.vptree_flat.vptree import VPTree as FVPT
-from tdamapper.utils.vptree_hier.vptree import VPTree as HVPT
+from tdamapper.metrics import euclidean
+from tdamapper.vptree_flat.vptree import VPTree as FVPT
+from tdamapper.vptree_hier.vptree import VPTree as HVPT
 from tests.ball_tree import SkBallTree
 from tests.setup_logging import setup_logging
 
diff --git a/tests/test_bench_metrics.py b/tests/test_bench_metrics.py
@@ -5,7 +5,7 @@
 import numpy as np
 import pandas as pd
 
-import tdamapper.utils.metrics as metrics
+import tdamapper.metrics as metrics
 from tests.setup_logging import setup_logging
 
 
diff --git a/tests/test_bench_vptree.py b/tests/test_bench_vptree.py
@@ -5,9 +5,9 @@
 from sklearn.datasets import load_breast_cancer, load_digits, load_iris
 
 from tdamapper._common import profile
-from tdamapper.utils.metrics import euclidean, get_metric
-from tdamapper.utils.vptree_flat.vptree import VPTree as FVPT
-from tdamapper.utils.vptree_hier.vptree import VPTree as HVPT
+from tdamapper.metrics import euclidean, get_metric
+from tdamapper.vptree_flat.vptree import VPTree as FVPT
+from tdamapper.vptree_hier.vptree import VPTree as HVPT
 from tests.ball_tree import SkBallTree
 from tests.setup_logging import setup_logging
 
diff --git a/tests/test_unit_heap.py b/tests/test_unit_heap.py
@@ -1,6 +1,6 @@
 import random
 
-from tdamapper.utils.heap import MaxHeap
+from tdamapper.heap import MaxHeap
 
 
 def maxheap(data):
diff --git a/tests/test_unit_knn.py b/tests/test_unit_knn.py
@@ -1,8 +1,8 @@
 import numpy as np
 
+from tdamapper.metrics import euclidean
 from tdamapper.search import KNNSearch
-from tdamapper.utils.metrics import euclidean
-from tdamapper.utils.vptree_flat.vptree import VPTree
+from tdamapper.vptree_flat.vptree import VPTree
 
 X = np.array(
     [
diff --git a/tests/test_unit_metrics.py b/tests/test_unit_metrics.py
@@ -1,6 +1,6 @@
 import numpy as np
 
-import tdamapper.utils.metrics as metrics
+import tdamapper.metrics as metrics
 
 
 def test_euclidean():
diff --git a/tests/test_unit_quickselect.py b/tests/test_unit_quickselect.py
@@ -2,7 +2,7 @@
 
 import numpy as np
 
-from tdamapper.utils.quickselect import partition, quickselect
+from tdamapper.quickselect import partition, quickselect
 
 
 def test_partition():
diff --git a/tests/test_unit_search.py b/tests/test_unit_search.py
@@ -0,0 +1,42 @@
+import numpy as np
+
+from tdamapper.search import BallSearch, CubicalSearch, KNNSearch
+
+
+def test_ball_search():
+    search = BallSearch(radius=1.0)
+    data = np.array([[0.0], [1.0], [2.0]])
+    search.fit(data)
+
+    assert 1 == len(search.search(np.array([0.0])))
+
+    assert 0 in search.search(np.array([0.0]))
+    assert 1 not in search.search(np.array([0.0]))
+
+    assert 0 in search.search(np.array([0.5]))
+    assert 1 in search.search(np.array([0.5]))
+
+
+def test_knn_search():
+    search = KNNSearch(neighbors=2)
+    data = np.array([[0.0], [1.0], [2.0], [3.0]])
+    search.fit(data)
+
+    assert 2 == len(search.search(np.array([0.0])))
+
+    assert 0 in search.search(np.array([0.0]))
+    assert 1 in search.search(np.array([0.0]))
+
+    assert 0 in search.search(np.array([0.5]))
+    assert 1 in search.search(np.array([0.5]))
+
+
+def test_cubical_search():
+    search = CubicalSearch(n_intervals=4, overlap_frac=0.25)
+    data = np.array([[0.0], [1.0], [2.0], [3.0], [4.0]])
+    search.fit(data)
+
+    assert 2 == len(search.search(np.array([1.0])))
+
+    assert 1 in search.search(np.array([1.0]))
+    assert 2 in search.search(np.array([1.0]))
diff --git a/tests/test_unit_unionfind.py b/tests/test_unit_unionfind.py
@@ -1,4 +1,4 @@
-from tdamapper.utils.unionfind import UnionFind
+from tdamapper.unionfind import UnionFind
 
 
 def test_list():
diff --git a/tests/test_unit_vptree.py b/tests/test_unit_vptree.py
@@ -2,9 +2,9 @@
 
 import numpy as np
 
-from tdamapper.utils.metrics import get_metric
-from tdamapper.utils.vptree_flat.vptree import VPTree as FVPT
-from tdamapper.utils.vptree_hier.vptree import VPTree as HVPT
+from tdamapper.metrics import get_metric
+from tdamapper.vptree_flat.vptree import VPTree as FVPT
+from tdamapper.vptree_hier.vptree import VPTree as HVPT
 from tests.ball_tree import SkBallTree
 
 distance = "euclidean"
