feat(browser): auto-commit con 178 cambios

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

python/functions/datascience/kmeans_segments_test.py

@@ -0,0 +1,64 @@
+"""Tests para kmeans_segments."""
+
+import numpy as np
+
+from kmeans_segments import kmeans_segments
+
+
+def _three_blobs(seed: int = 0, per_blob: int = 40):
+    """Genera 3 blobs gaussianos bien separados en 2D, alineados por fila."""
+    rng = np.random.default_rng(seed)
+    centers = [(0.0, 0.0), (12.0, 12.0), (0.0, 12.0)]
+    xs: list[float] = []
+    ys: list[float] = []
+    for cx, cy in centers:
+        pts = rng.normal(loc=(cx, cy), scale=0.4, size=(per_blob, 2))
+        xs.extend(float(p[0]) for p in pts)
+        ys.extend(float(p[1]) for p in pts)
+    return {"x": xs, "y": ys}
+
+
+def test_three_separated_blobs_finds_k3():
+    columns = _three_blobs(seed=0, per_blob=40)
+    result = kmeans_segments(columns, k_min=2, k_max=8)
+
+    assert result["best_k"] == 3
+    assert result["silhouette"] > 0.5
+    assert result["n_features"] == 2
+    assert result["n_rows_used"] == 120
+    assert sum(result["cluster_sizes"]) == 120
+    assert len(result["centers"]) == 3
+    # scores_by_k cubre todo el rango probado.
+    ks = [s["k"] for s in result["scores_by_k"]]
+    assert ks == list(range(2, 9))
+
+
+def test_insufficient_rows_returns_note():
+    # Solo 3 filas válidas, k_min*2 = 4 -> insuficiente.
+    columns = {"x": [1.0, 2.0, 3.0], "y": [1.0, 2.0, 3.0]}
+    result = kmeans_segments(columns, k_min=2, k_max=8)
+
+    assert result["best_k"] == 0
+    assert result["note"] == "datos insuficientes"
+
+
+def test_insufficient_numeric_columns_returns_note():
+    # Una sola columna numérica; la otra es texto -> menos de 2 numéricas.
+    columns = {
+        "x": [1.0, 2.0, 3.0, 4.0, 5.0, 6.0],
+        "label": ["a", "b", "c", "d", "e", "f"],
+    }
+    result = kmeans_segments(columns, k_min=2, k_max=8)
+
+    assert result["best_k"] == 0
+    assert result["note"] == "datos insuficientes"
+
+
+def test_rows_with_none_are_dropped():
+    columns = _three_blobs(seed=1, per_blob=40)
+    # Inyectar None en una fila; debe descartarse, dejando 119.
+    columns["x"][0] = None
+    result = kmeans_segments(columns, k_min=2, k_max=8)
+
+    assert result["best_k"] == 3
+    assert result["n_rows_used"] == 119