Merge pull request #509 from sjsrey/encoding

Remove sklearn dependency

Author: sjsrey

libpysal/weights/tests/test_weights.py

@@ -3,7 +3,7 @@
 
 import unittest
 import pytest
-from ..weights import W, WSP
+from ..weights import W, WSP, _LabelEncoder
 from .. import util
 from ..util import WSP2W, lat2W
 from ..contiguity import Rook
@@ -724,6 +724,15 @@ def test_from_WSP(self):
         self.assertEqual(w.n, 100)
         self.assertEqual(w.pct_nonzero, 4.62)
 
+    def test_LabelEncoder(self):
+        le = _LabelEncoder()
+        le.fit(["NY", "CA", "NY", "CA", "TX", "TX"])
+        np.testing.assert_equal(le.classes_, np.array(["CA", "NY", "TX"]))
+        np.testing.assert_equal(
+            le.transform(["NY", "CA", "NY", "CA", "TX", "TX"]),
+            np.array([1, 0, 1, 0, 2, 2]),
+        )
+
 
 if __name__ == "__main__":
     unittest.main()

libpysal/weights/weights.py

@@ -10,9 +10,9 @@
 import numpy as np
 import scipy.sparse
 from scipy.sparse.csgraph import connected_components
-from sklearn import preprocessing
 from collections import defaultdict
 
+
 # from .util import full, WSP2W resolve import cycle by
 # forcing these into methods
 from . import adjtools
@@ -21,6 +21,56 @@
 __all__ = ["W", "WSP"]
 
 
+class _LabelEncoder(object):
+    """Encode labels with values between 0 and n_classes-1.
+
+    Attributes
+    ----------
+    classes_: array of shape [n_classes]
+        Class labels for each index.
+
+    Examples
+    --------
+    >>> le = _LabelEncoder()
+    >>> le.fit(["NY", "CA", "NY", "CA", "TX", "TX"])
+    >>> le.classes_
+    array(['CA', 'NY', 'TX'])
+    >>> le.transform(["NY", "CA", "NY", "CA", "TX", "TX"])
+    array([1, 0, 1, 0, 2, 2])
+    """
+
+    def fit(self, y):
+        """Fit label encoder.
+
+        Parameters
+        ----------
+        y : list
+            list of labels
+
+        Returns
+        -------
+        self : instance of self.
+          Fitted label encoder.
+        """
+        self.classes_ = np.unique(y)
+        return self
+
+    def transform(self, y):
+        """Transform labels to normalized encoding.
+
+        Parameters
+        ----------
+        y : list
+            list of labels
+
+        Returns
+        -------
+        y : array
+            array of normalized labels.
+        """
+        return np.searchsorted(self.classes_, y)
+
+
 class W(object):
     """
     Spatial weights class. Class attributes are described by their
@@ -505,7 +555,7 @@ def to_sparse(self, fmt="coo"):
             data = adj_list.weight
             row = adj_list.focal
             col = adj_list.neighbor
-            le = preprocessing.LabelEncoder()
+            le = _LabelEncoder()
             le.fit(row)
             row = le.transform(row)
             col = le.transform(col)