BUG: ProbabilisticPCA.score work with pipeline

score(X, y=None) is required to be used by generic objects that work on
supervised and unsupervised learners.

sklearn/decomposition/pca.py

@@ -320,7 +320,7 @@ class ProbabilisticPCA(PCA):
             self.covariance_ += self.explained_variance_[k] * add_cov
         return self
 
-    def score(self, X):
+    def score(self, X, y=None):
         """Return a score associated to new data
 
         Parameters