[Bug] Configs with different types (numpy vs native) are counted double

CHANGELOG.md

@@ -1,3 +1,8 @@
+# 2.3.2
+
+## Bugfixes
+- Fix that config could be accidentally added twice to the runhistory (#1242)
+
 # 2.3.1
 
 ## Bugfixes

smac/intensifier/successive_halving.py

@@ -209,7 +209,7 @@ def get_state(self) -> dict[str, Any]:  # noqa: D102
         for key in list(self._tracker.keys()):
             for seed, configs in self._tracker[key]:
                 # We have to make key serializable
-                new_key = f"{key[0]},{key[1]}"
+                new_key = f"{key[0]},{key[1]}"  # noqa: E231
                 tracker[new_key].append((seed, [dict(config) for config in configs]))
 
         return {"tracker": tracker}
@@ -260,7 +260,7 @@ def print_tracker(self) -> None:
                 messages.append(f"--- Bracket {bracket} / Stage {stage}: {counter} configs")
 
         if len(messages) > 0:
-            logger.debug(f"{self.__class__.__name__} statistics:")
+            logger.debug(f"{self.__class__.__name__} statistics:")  # noqa: E231
 
         for message in messages:
             logger.debug(message)

smac/main/smbo.py

@@ -325,8 +325,8 @@ def optimize(self, *, data_to_scatter: dict[str, Any] | None = None) -> Configur
 
             # Some statistics
             logger.debug(
-                f"Remaining wallclock time: {self.remaining_walltime}; "
-                f"Remaining cpu time: {self.remaining_cputime}; "
+                f"Remaining wallclock time: {self.remaining_walltime}; "  # noqa: E702
+                f"Remaining cpu time: {self.remaining_cputime}; "  # noqa: E702
                 f"Remaining trials: {self.remaining_trials}"
             )
 
@@ -538,7 +538,7 @@ def _initialize_state(self) -> None:
                     )
                     logger.info(
                         f"Found old run in `{self._scenario.output_directory}` but it is not the same as the current "
-                        f"one:\n{diff}"
+                        f"one:\n{diff}"  # noqa: E231
                     )
 
                     feedback = input(

smac/runhistory/runhistory.py

@@ -167,6 +167,25 @@ def empty(self) -> bool:
         """
         return len(self._data) == 0
 
+    def _convert_config_to_pure_python(self, config: Configuration) -> Configuration:
+        """Convert config values from possibly numpy types to python built ins.
+
+        Background: The configs are perceived as different when they have a np.int type instead of int.
+
+        Args:
+            config (Configuration): Config to convert to pure python.
+
+        Returns
+        -------
+            Configuration
+                Config with only pure python types.
+        """
+        config_values = list(dict(config).values())
+        config_values = json.loads(json.dumps(config_values, cls=NumpyEncoder))
+        config_keys = list(dict(config).keys())
+        config = Configuration(configuration_space=config.config_space, values=dict(zip(config_keys, config_values)))
+        return config
+
     def add(
         self,
         config: Configuration,
@@ -211,6 +230,8 @@ def add(
         if additional_info is None:
             additional_info = {}
 
+        config = self._convert_config_to_pure_python(config)
+
         # Squeeze is important to reduce arrays with one element
         # to scalars.
         cost_array = np.asarray(cost).squeeze()

smac/utils/data_structures.py

@@ -51,7 +51,7 @@ def recursively_compare_dicts(
 
     elif isinstance(d1, list) and isinstance(d2, list):
         if len(d1) != len(d2):
-            diff += [f"{level}: len1={len(d1)}; len2={len(d2)}"]
+            diff += [f"{level}: len1={len(d1)}; len2={len(d2)}"]  # noqa: E702
             # logger.info("{:<20} len1={}; len2={}".format(level, len(d1), len(d2)))
             # logger.info("len1={}; len2={}".format(len(d1), len(d2)))
         common_len = min(len(d1), len(d2))