Update env.py

env.py

@@ -153,11 +153,11 @@ class TrafficEnv:
             "avg_waiting_time":     0.0,
             "max_queue_length":     0,
             "signal_switch_count":  0,
-            "congestion_score":     0.0,
+            "congestion_score":     1e-5,
             "avg_ev_clear_time":    0.0,
             "total_ev_cleared":     0,
             "total_ev_penalty":     0.0,
-            "fairness_score":       1.0,
+            "fairness_score":       1.0 - 1e-5,
         }
 
         # Track waiting steps for emergency vehicles and phase stability
@@ -247,9 +247,10 @@ class TrafficEnv:
         total_wait_sum = sum(self.waiting_times.values())
         denom = max(1, self.total_cleared)
         self._metrics["avg_waiting_time"] = total_wait_sum / denom
-        self._metrics["congestion_score"] = (
-            sum(self.queues.values()) / (self.max_queue * len(LANES))
-        )
+        self._metrics["congestion_score"] = float(np.clip(
+            sum(self.queues.values()) / (self.max_queue * len(LANES)),
+            1e-5, 1.0 - 1e-5
+        ))
 
         # ── 7. Calculate reward ───────────────────────────────────────
         post_total_queue = sum(self.queues.values())
@@ -265,7 +266,12 @@ class TrafficEnv:
         # Simple fairness: (1 - variance of wait times / threshold)
         wait_vals = list(self.waiting_times.values())
         if max(wait_vals) > 0:
-            self._metrics["fairness_score"] = max(0.0, 1.0 - (np.std(wait_vals) / self.starvation_limit))
+            self._metrics["fairness_score"] = float(np.clip(
+                1.0 - (np.std(wait_vals) / self.starvation_limit),
+                1e-5, 1.0 - 1e-5
+            ))
+        else:
+            self._metrics["fairness_score"] = 1.0 - 1e-5
 
         # ── 9. Termination ────────────────────────────────────────────
         done = self.step_count >= self.max_steps