better coloration

sim.py

@@ -10,11 +10,11 @@ class Direction(Enum):
     EAST = (1, 0)
 
 class Colours(Enum):
-    WALL = [30,30,30]
-    STOCKED = [200,150,0]
-    DEPLETED = [100,40,0]
-    UNDEFINED = [255,0,255]
-    AGENT = [0,255,0]
+    WALL = [30, 30, 30]
+    STOCKED = [240, 170, 0]
+    DEPLETED = [77, 61, 41]
+    UNDEFINED = [255, 0, 255]
+    AGENT = [0, 255, 0]
 
 @dataclass
 class Observations:
@@ -53,9 +53,6 @@ class Environment:
             )),
             np.ones((shape[0], 1))
         ))
-        
-        print(self.obstacle_map)
-        print(self.obstacle_map.shape)
     
     def step(self):
         agent_map = np.zeros(self.shape)
@@ -74,16 +71,31 @@ class Environment:
     def render(self):
         pixel_data = np.zeros(self.shape + (3,))
 
-        print(pixel_data.shape)
-
         pixel_data += (
-            ((self.obstacle_map[..., np.newaxis] == 1) * np.array(Colours.WALL.value))
-            + ((self.obstacle_map[..., np.newaxis] == 0) * (self.resource_map[..., np.newaxis] == 0) * np.array(Colours.DEPLETED.value))
-            + ((self.obstacle_map[..., np.newaxis] == 0) * (self.resource_map[..., np.newaxis] > 0) * np.array(Colours.STOCKED.value))
+            (
+                (self.obstacle_map[..., np.newaxis] == 1)   # Flag: wall 
+                * np.array(Colours.WALL.value)              # Colour
+            )
+            + (
+                (self.obstacle_map[..., np.newaxis] == 0)   # Flag: no wall
+                * (self.resource_map[..., np.newaxis] == 0) # Flag: depleted
+                * np.array(Colours.DEPLETED.value)          # Colour
+            )
+            + (
+                (
+                    # Set of flags for applying the "cell is stocked" colour component
+                    (self.obstacle_map[..., np.newaxis] == 0)  # Flag: no wall
+                    * (self.resource_map[..., np.newaxis] > 0) # Flag: not depleted
+                ) * (
+                    # Remapping of resource values to visualize relative cell resources
+                    self.resource_map[..., np.newaxis] # Resource values
+                    / 100 # Remap of resource values to (vaguely) 0-1 float range
+                    * 0.4 # Remap of 0-1 float range to 0-0.2 float range
+                    + 0.6 # Raising baseline of 0-0.2 float range to 0.8-1 float range
+                ) 
+                * np.array(Colours.STOCKED.value)          # Colour
+            ) 
         )
-
-        print(pixel_data)
-        print(pixel_data.shape)
         
         for agent in self.agents:
             pixel_data[tuple(agent.position)] = agent.get_colour()
@@ -210,10 +222,7 @@ class CAMSReverseAndSidestepAgent(DirectionalAgent):
                 self.move_forward()
                 self.number_of_turns += 1
             elif self.number_of_turns > 0:
-                self.turn_right()
-                self.move_forward()
-                self.turn_right()
-                self.number_of_turns = 0
+                self.sidestep_and_reset_counter()
         elif self.resources >= self.required_resources:
             self.resources -= self.required_resources
             new_agent = CAMSReverseAndSidestepAgent(
@@ -226,13 +235,17 @@ class CAMSReverseAndSidestepAgent(DirectionalAgent):
                 self.reverse_direction()
                 self.number_of_turns += 1
             elif self.number_of_turns > 0:
-                self.turn_right()
-                self.move_forward()
-                self.turn_right()
-                self.number_of_turns = 0
+                self.sidestep_and_reset_counter()
         else:
             self.move_forward()
     
+    def sidestep_and_reset_counter(self):
+        self.turn_right()
+        self.move_forward()
+        self.turn_right()
+        self.number_of_turns = 0
+
+
     def eat(self):
         resources = self.environment.eat(self.position)