Update from vcha/stable

pufferlib/config/ocean/drive.ini

@@ -124,6 +124,8 @@ traffic_control_scope = 0
 ; Erratic agents are selected once per episode; masked out of the PPO rollout buffer.
 ; Per-episode probability that an agent is blind to other agents for the whole episode
 partner_blindness_prob = 0.0
+; Per-step probability that a flagged blind-partner agent zeros out its partner observations
+partner_blindness_trigger_prob = 0.0
 ; Per-episode probability that an agent is flagged as a phantom braker for the whole episode
 phantom_braking_prob = 0.0
 ; Per-step probability that a flagged phantom-braker actually initiates a brake window

pufferlib/ocean/drive/binding.c

@@ -1846,6 +1846,7 @@ static int my_init(Env *env, PyObject *args, PyObject *kwargs) {
     env->obs_lane_segment_count = (int) unpack(kwargs, "obs_lane_segment_count");
     env->obs_boundary_segment_count = (int) unpack(kwargs, "obs_boundary_segment_count");
     env->partner_blindness_prob = (float) unpack(kwargs, "partner_blindness_prob");
+    env->partner_blindness_trigger_prob = (float) unpack(kwargs, "partner_blindness_trigger_prob");
     env->phantom_braking_prob = (float) unpack(kwargs, "phantom_braking_prob");
     env->phantom_braking_trigger_prob = (float) unpack(kwargs, "phantom_braking_trigger_prob");
     env->phantom_braking_duration = (int) unpack(kwargs, "phantom_braking_duration");
@@ -1895,7 +1896,6 @@ static int my_log(PyObject *dict, Env *env, Log *log, float n) {
         // Puffer score components
         assign_to_dict(dict, "at_fault_collision_rate", log->at_fault_collision_rate);
         assign_to_dict(dict, "puffer_score", log->puffer_score);
-        assign_to_dict(dict, "ttc_within_bound_rate", log->ttc_within_bound_rate);
         assign_to_dict(dict, "driving_direction_score", log->driving_direction_score);
         assign_to_dict(dict, "speed_limit_compliance", log->speed_limit_compliance);
         assign_to_dict(dict, "making_progress_rate", log->making_progress_rate);

pufferlib/ocean/drive/datatypes.h

@@ -1,3 +1,5 @@
+#include <stdlib.h>
+
 // -- REWARD CONDITIONING COEFFICIENTS
 #define REWARD_COEF_GOAL_RADIUS 0
 #define REWARD_COEF_GOAL_SPEED 1

pufferlib/ocean/drive/drive.c

@@ -127,6 +127,7 @@ void demo() {
         .max_traffic_control_observations = conf.max_traffic_control_observations,
         .traffic_control_scope = conf.traffic_control_scope,
         .partner_blindness_prob = conf.partner_blindness_prob,
+        .partner_blindness_trigger_prob = conf.partner_blindness_trigger_prob,
         .phantom_braking_prob = conf.phantom_braking_prob,
         .phantom_braking_trigger_prob = conf.phantom_braking_trigger_prob,
         .phantom_braking_duration = conf.phantom_braking_duration,
@@ -251,6 +252,7 @@ void performance_test() {
         .max_traffic_control_observations = conf.max_traffic_control_observations,
         .traffic_control_scope = conf.traffic_control_scope,
         .partner_blindness_prob = conf.partner_blindness_prob,
+        .partner_blindness_trigger_prob = conf.partner_blindness_trigger_prob,
         .phantom_braking_prob = conf.phantom_braking_prob,
         .phantom_braking_trigger_prob = conf.phantom_braking_trigger_prob,
         .phantom_braking_duration = conf.phantom_braking_duration,