Added catch for predictors that do not support summary()

src/python/nimbusml/pipeline.py

@@ -2543,7 +2543,7 @@ def summary(self, verbose=0, **params):
         if len(self.steps) > 0 and not isinstance(
                 self.last_node, BasePredictor):
             raise ValueError(
-                "Summary is availabe only for predictor types, instead "
+                "Summary is available only for predictor types, instead "
                 "got " +
                 self.last_node.type)
 
@@ -2581,6 +2581,10 @@ def summary(self, verbose=0, **params):
             self._run_time = time.time() - start_time
             raise e
 
+        # .summary() not supported if size of summary_data
+        # is less or equal to 1 (if only PredictedName in summary_data)
+        if summary_data.size == 1 and summary_data.columns.values == ["PredictorName"]:
+            raise TypeError("One or more predictors in this pipeline do not support the .summary() function.")
         self.model_summary = summary_data
 
         # stop the clock

src/python/nimbusml/tests/model_summary/test_model_summary.py

@@ -66,25 +66,25 @@
     #SymSgdBinaryClassifier(),
     OrdinaryLeastSquaresRegressor(),
     PoissonRegressionRegressor(),
-    OneVsRestClassifier(FastLinearBinaryClassifier()),
     GamRegressor(),
     GamBinaryClassifier(),
     PcaAnomalyDetector(),
-    FactorizationMachineBinaryClassifier(),
-    KMeansPlusPlus(n_clusters=2),
-    NaiveBayesClassifier(),
     FastForestBinaryClassifier(number_of_trees=2), 
     FastForestRegressor(number_of_trees=2),
     FastTreesBinaryClassifier(number_of_trees=2),
     FastTreesRegressor(number_of_trees=2),
     FastTreesTweedieRegressor(number_of_trees=2),
     LightGbmRegressor(number_of_iterations=2),
-    LightGbmClassifier(),
     LightGbmBinaryClassifier(number_of_iterations=2)
 ]
 
 learners_not_supported = [
-    #PcaTransformer(), # REVIEW: crashes
+    FactorizationMachineBinaryClassifier(),
+    OneVsRestClassifier(FastLinearBinaryClassifier()),
+    FactorizationMachineBinaryClassifier(),
+    KMeansPlusPlus(n_clusters=2),
+    NaiveBayesClassifier(),
+    LightGbmClassifier()
 ]
 
 
@@ -98,7 +98,6 @@ def test_model_summary(self):
             pipeline.fit(train_stream, label_column)
             pipeline.summary()
 
-    @unittest.skip("No unsupported learners")
     def test_model_summary_not_supported(self):
         for learner in learners_not_supported:
             pipeline = Pipeline(
@@ -107,6 +106,23 @@ def test_model_summary_not_supported(self):
             pipeline.fit(train_stream, label_column)
             assert_raises(TypeError, pipeline.summary)
 
+    def test_model_summary_not_supported_specific(self):
+        path = get_dataset('infert').as_filepath()
+        data = FileDataStream.read_csv(path, sep=',',
+                               names={0: 'row_num', 5: 'case'})
+        pipeline = Pipeline([
+            OneHotVectorizer(columns={'edu': 'education'}),
+            FactorizationMachineBinaryClassifier(feature=['induced', 'edu', 'parity'],
+                                         label='case')
+        ])
+        pipeline.fit(data)
+        try:
+            pipeline.summary()
+        except TypeError as e:
+            self.assertEqual(e.args[0], "One or more predictors in this pipeline do not support the .summary() function.")
+        else:
+            assert False
+
     def test_summary_called_back_to_back_on_predictor(self):
         """
         When a predictor is fit without using a Pipeline,