Delete the cached summaries when refitting a pipeline or a predictor. (…

…#109)

* Fix build issue on Windows when VS2019 is installed.

Note: The -version option could not be added directly
to the FOR command due to a command script parsing issue.

* Add missing arguments to fix build issue with latest version of autoflake.

* Delete the cached summaries when refitting a pipeline or a predictor.
Fixes #106

* Simplify the code that deletes cached summaries when calling fit.

src/python/nimbusml/base_predictor.py

@@ -52,6 +52,11 @@ def fit(self, X, y=None, **params):
         self.X_ = X
         self.y_ = y
 
+        # Clear cached summary since it should not
+        # retain its value after a new call to fit
+        if hasattr(self, 'model_summary_'):
+            delattr(self, 'model_summary_')
+
         pipeline = Pipeline([self])
         try:
             pipeline.fit(X, y, **params)

src/python/nimbusml/pipeline.py

@@ -1089,13 +1089,15 @@ def fit(self, X, y=None, verbose=1, **params):
                 clone = self.clone()
             self.steps = clone.steps
 
+        # Clear cached values
+        for attr in ["_run_time_error", "model_summary"]:
+            if hasattr(self, attr):
+                delattr(self, attr)
+
         # Caches the predictor to restore it as it was
         # in case of exception. It is deleted after the training.
         self._cache_predictor = deepcopy(self.steps[-1])
 
-        if hasattr(self, "_run_time_error"):
-            delattr(self, "_run_time_error")
-
         # Checks that no node was ever trained.
         for i, n in enumerate(self.nodes):
             if hasattr(n, "model_") and n.model_ is not None:

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

@@ -123,6 +123,29 @@ def test_summary_called_back_to_back_on_predictor(self):
         ols.summary()
         ols.summary()
 
+    def test_pipeline_summary_is_refreshed_after_refitting(self):
+        predictor = OrdinaryLeastSquaresRegressor(normalize='No', l2_regularization=0)
+        pipeline = Pipeline([predictor])
+
+        pipeline.fit([0,1,2,3], [1,2,3,4])
+        summary1 = pipeline.summary()
+
+        pipeline.fit([0,1,2,3], [2,5,8,11])
+        summary2 = pipeline.summary()
+
+        self.assertFalse(summary1.equals(summary2))
+
+    def test_predictor_summary_is_refreshed_after_refitting(self):
+        predictor = OrdinaryLeastSquaresRegressor(normalize='No', l2_regularization=0)
+
+        predictor.fit([0,1,2,3], [1,2,3,4])
+        summary1 = predictor.summary()
+
+        predictor.fit([0,1,2,3], [2,5,8,11])
+        summary2 = predictor.summary()
+
+        self.assertFalse(summary1.equals(summary2))
+
 
 if __name__ == '__main__':
     unittest.main()