A TODO added to controller_base.py.

backends/controller_base.py

@@ -52,6 +52,8 @@ class ControllerBase(PolicyBase):
     Returns state at end of node execution, total reward, epsiode_termination_flag, info
     '''
     # TODO: this is never called when you TEST high-level policy (w/o MCTS) rather than train...
+    # (make some integrated interface btw testing and training and b.t.w. the high- and low-level
+    # methods with and without MCTS.
     def step_current_node(self, visualize_low_level_steps=False):
         total_reward = 0
         self.node_terminal_state_reached = False