Update the developer docs.

doc/developers/index.rst

@@ -219,9 +219,8 @@ accept as arguments constants that determine the estimator behavior
 It should not, however, take the actual training data as argument, as
 this is leaved to the ``fit()`` method::
 
-    clf1 = SVM(impl='c_svm')
-    clf2 = SVM(C=2.3)
-    clf3 = SVM([[1, 2], [2, 3]], [-1, 1]) # WRONG!
+    clf2 = SVC(C=2.3)
+    clf3 = SVC([[1, 2], [2, 3]], [-1, 1]) # WRONG!
 
 
 The arguments that go in the `__init__` should all be keyword arguments
@@ -262,6 +261,8 @@ data in the predict method.
       samples and D is the number of features.
     * Y : array, with shape = [N], where N is the number of samples.
 
+    * args, kwargs. Parameters can also be set in the fit method.
+
 X.shape[0] should be the same as Y.shape[0]. If this requisite is not
 met, an exception should be raised.