Refactor/testing logic (#140)

* disable testing if test data is not provided

* change message

Author: voorhs

autointent/_pipeline/_pipeline.py

@@ -141,6 +141,13 @@ def fit(
         context.configure_vector_index(self.vector_index_config)
 
         self.validate_modules(dataset)
+
+        test_utterances = context.data_handler.test_utterances()
+        if test_utterances is None:
+            self._logger.warning(
+                "Test data is not provided. Final test metrics won't be calculated after pipeline optimization."
+            )
+
         self._fit(context, sampler)
 
         if context.is_ram_to_clear():
@@ -153,15 +160,17 @@ def fit(
         self.nodes = {node.node_type: node for node in nodes_list}
 
         if refit_after:
+            # TODO reflect this refitting in dumped version of pipeline
             self._refit(context)
 
-        predictions = self.predict(context.data_handler.test_utterances())
-        for metric_name, metric in DECISION_METRICS.items():
-            context.optimization_info.pipeline_metrics[metric_name] = metric(
-                context.data_handler.test_labels(),
-                predictions,
-            )
-        context.callback_handler.log_final_metrics(context.optimization_info.pipeline_metrics)
+        if test_utterances is not None:
+            predictions = self.predict(test_utterances)
+            for metric_name, metric in DECISION_METRICS.items():
+                context.optimization_info.pipeline_metrics[metric_name] = metric(
+                    context.data_handler.test_labels(),
+                    predictions,
+                )
+            context.callback_handler.log_final_metrics(context.optimization_info.pipeline_metrics)
 
         return context
 

autointent/context/data_handler/_data_handler.py

@@ -150,7 +150,7 @@ def validation_labels(self, idx: int | None = None) -> ListOfGenericLabels:
         split = self._choose_split(Split.VALIDATION, idx)
         return cast(ListOfGenericLabels, self.dataset[split][self.dataset.label_feature])
 
-    def test_utterances(self) -> list[str]:
+    def test_utterances(self) -> list[str] | None:
         """
         Retrieve test utterances from the dataset.
 
@@ -161,6 +161,8 @@ def test_utterances(self) -> list[str]:
         :param idx: Optional index for a specific test split.
         :return: List of test utterances.
         """
+        if Split.TEST not in self.dataset:
+            return None
         return cast(list[str], self.dataset[Split.TEST][self.dataset.utterance_feature])
 
     def test_labels(self) -> ListOfGenericLabels: