canonical example working with new Dedupe api class

dedupeio · Sep 13, 2012 · 8a8ecbe · 8a8ecbe
1 parent ae38870
commit 8a8ecbe
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 6 deletions.
diff --git a/dedupe/dedupe.py b/dedupe/dedupe.py
@@ -157,15 +157,15 @@ def initializeTraining(self, training_file=None) :
         if training_file :
             (self.training_pairs,
              self.training_data) = self.readTraining(training_source,
-                                                     training_data)                    
+                                                     self.training_data)                    
 
 
 
-    def trainX(self, training_source=None, data_d) :
-
-        elif (training_source.__class__ is not str
-              or not isinstance(training_source, types.FunctionType):
+    def trainX(self, data_d, training_source=None) :
+        if (training_source.__class__ is not str
+              or not isinstance(training_source, types.FunctionType)):
             raise ValueError
+
         self.data_d = sampleDict(data_d, 700)
 
         if training_source.__class__ is str:

diff --git a/examples/canonical_example.py b/examples/canonical_example.py
@@ -89,7 +89,10 @@ def printPairs(pairs):
                                                    num_training_dupes,
                                                    num_training_distinct)
 
-    deduper.trainingDistance()
+    deduper.training_data = dedupe.training_sample.addTrainingData(deduper.training_pairs,
+                                                            deduper.data_model,
+                                                            deduper.training_data)
+
     deduper.train()
 
 deduper.findDuplicates(data_d)