BUG: Make SVMs work on non contiguous arrays

GaelVaroquaux · Fabian Pedregosa · commit 847d00203779 · 2011-07-01T09:10:54.000+02:00
diff --git a/doc/modules/svm.rst b/doc/modules/svm.rst
@@ -37,6 +37,13 @@ The disadvantages of Support Vector Machines include:
 
 .. TODO: add reference to probability estimates
 
+.. note:: **Avoiding data copy**
+
+    If the data passed to certain methods is not C-ordered and
+    contiguous, it will be copied before calling the underlying C
+    implementation. You can check whether a give numpy array is
+    C-contiguous by inspecting its `flags` dictionnary.
+
 .. _svm_classification:
 
 Classification
diff --git a/scikits/learn/svm/base.py b/scikits/learn/svm/base.py
@@ -99,6 +99,12 @@ class frequencies.
         -------
         self : object
             Returns self.
+
+        Notes
+        ------
+        If X and y are not C-ordered and contiguous arrays, they are
+        copied.
+
         """
         self._set_params(**params)
 
@@ -133,7 +139,8 @@ class frequencies.
         self.support_, self.support_vectors_, self.n_support_, \
         self.dual_coef_, self.intercept_, self.label_, self.probA_, \
         self.probB_ = libsvm.fit(
-            X, y, svm_type=solver_type, sample_weight=sample_weight,
+            np.ascontiguousarray(X), np.ascontiguousarray(y), 
+            svm_type=solver_type, sample_weight=sample_weight,
             class_weight=class_weight,
             class_weight_label=class_weight_label,
             **self._get_params())
diff --git a/scikits/learn/svm/classes.py b/scikits/learn/svm/classes.py
@@ -136,7 +136,6 @@ class SVC(BaseLibSVM, ClassifierMixin):
     `intercept_` : array, shape = [n_class * (n_class-1) / 2]
         Constants in decision function.
 
-
     Examples
     --------
     >>> import numpy as np
@@ -575,6 +574,11 @@ def fit(self, X, class_weight={}, sample_weight=[], **params):
         -------
         self : object
             Returns self.
+
+        Notes
+        ------
+        If X is not a C-ordered contiguous array, it is copied.
+
         """
         super(OneClassSVM, self).fit(
             X, [], class_weight=class_weight, sample_weight=sample_weight,
