Update run_indefinetly to directly inherit from TransitionCriterion (f…

…acebookresearch#329)

Summary:

This diff we update the 4 of 4 aepsych completion criterion to directly rely on transitioncriterion

In following diffs we will:
- Completely remove the completion criterion file
- revisit storage
- remove all todos in gennode, genstrat, and transitioncriterion classes related to maintaining this deprecated code
- update AEPsych GSs as needed
- determine if run indefinetly can be replaced by simply having gen_unlimited_trials = true
- determine if additional updates to min_asks and run_indefinetly are necessary

Reviewed By: crasanders

Differential Revision: D52853023

aepsych/generators/completion_criterion/run_indefinitely.py

@@ -5,20 +5,29 @@
 # This source code is licensed under the license found in the
 # LICENSE file in the root directory of this source tree.
 
-from typing import Any, Dict
+from typing import Any, Dict, Optional, Set
 
 from aepsych.config import Config, ConfigurableMixin
 from ax.core import Experiment
-from ax.modelbridge.completion_criterion import CompletionCriterion
+from ax.modelbridge.transition_criterion import TransitionCriterion
 
 
-class RunIndefinitely(CompletionCriterion, ConfigurableMixin):
+class RunIndefinitely(TransitionCriterion, ConfigurableMixin):
     def __init__(self, run_indefinitely: bool) -> None:
         self.run_indefinitely = run_indefinitely
 
     def is_met(self, experiment: Experiment) -> bool:
         return not self.run_indefinitely
 
+    def block_continued_generation_error(
+        self,
+        node_name: Optional[str],
+        model_name: Optional[str],
+        experiment: Optional[Experiment],
+        trials_from_node: Optional[Set[int]] = None,
+    ) -> None:
+        pass
+
     @classmethod
     def get_config_options(cls, config: Config, name: str) -> Dict[str, Any]:
         run_indefinitely = config.getboolean(name, "run_indefinitely", fallback=False)