use explicit b'' literals for python3 compat

main.py

@@ -39,7 +39,7 @@ def _valid_target(target):
     return (
         0 <= target[0] < MAZE.shape[0] and
         0 <= target[1] < MAZE.shape[1] and
-        MAZE[tuple(target)] != '1'
+        MAZE[tuple(target)] != b'1'
     )
 
 
@@ -52,7 +52,7 @@ def init_global(maze_filename):
         maze_filename,
         dtype='|S1',
     )
-    state_mask = (MAZE != '1')
+    state_mask = (MAZE != b'1')
 
     S_TO_IJ = np.indices(MAZE.shape).transpose(1, 2, 0)[state_mask]
     SN = len(S_TO_IJ)
@@ -62,12 +62,12 @@ def init_global(maze_filename):
 
     # One step cost functions initialization
     maze_cost = np.zeros(MAZE.shape, dtype=np.float64)
-    maze_cost[MAZE == '1'] = np.nan
-    maze_cost[(MAZE == '0') | (MAZE == 'S')] = 0
-    maze_cost[MAZE == 'T'] = 50
-    maze_cost[MAZE == 'G'] = -1
+    maze_cost[MAZE == b'1'] = np.nan
+    maze_cost[(MAZE == b'0') | (MAZE == b'S')] = 0
+    maze_cost[MAZE == b'T'] = 50
+    maze_cost[MAZE == b'G'] = -1
     G1_X = maze_cost.copy()[state_mask]
-    maze_cost[(MAZE=='0') | (MAZE=='S') | (MAZE=='G')] += 1
+    maze_cost[(MAZE==b'0') | (MAZE==b'S') | (MAZE==b'G')] += 1
     G2_X = maze_cost.copy()[state_mask]
 
     # Actual environment modelling