Initial contextual breakout; exposed more variables

Author: Joseph Suarez

pufferlib/config/ocean/breakout.ini

@@ -10,7 +10,21 @@ num_envs = 8
 [env]
 num_envs = 1024
 frameskip = 4
-
+width = 576
+height = 330
+paddle_width = 62
+paddle_height = 8
+ball_width = 32
+ball_height = 32
+brick_width = 32
+brick_height = 12
+brick_rows = 6
+brick_cols = 18
+initial_ball_speed = 256
+max_ball_speed = 448
+paddle_speed = 620
+continuous = 0
+ 
 [policy]
 hidden_size = 128
 

pufferlib/ocean/breakout/binding.c

@@ -7,14 +7,17 @@ static int my_init(Env* env, PyObject* args, PyObject* kwargs) {
     env->frameskip = unpack(kwargs, "frameskip");
     env->width = unpack(kwargs, "width");
     env->height = unpack(kwargs, "height");
-    env->paddle_width = unpack(kwargs, "paddle_width");
+    env->initial_paddle_width = unpack(kwargs, "paddle_width");
     env->paddle_height = unpack(kwargs, "paddle_height");
     env->ball_width = unpack(kwargs, "ball_width");
     env->ball_height = unpack(kwargs, "ball_height");
     env->brick_width = unpack(kwargs, "brick_width");
     env->brick_height = unpack(kwargs, "brick_height");
     env->brick_rows = unpack(kwargs, "brick_rows");
     env->brick_cols = unpack(kwargs, "brick_cols");
+    env->initial_ball_speed = unpack(kwargs, "initial_ball_speed");
+    env->max_ball_speed = unpack(kwargs, "max_ball_speed");
+    env->paddle_speed = unpack(kwargs, "paddle_speed");
     env->continuous = unpack(kwargs, "continuous");
     init(env);
     return 0;

pufferlib/ocean/breakout/breakout.c

@@ -19,6 +19,9 @@ void demo() {
         .brick_height = 12,
         .brick_rows = 6,
         .brick_cols = 18,
+        .initial_ball_speed = 256,
+        .max_ball_speed = 448,
+        .paddle_speed = 620,
         .continuous = 0,
     };
     allocate(&env);

pufferlib/ocean/breakout/breakout.h

@@ -9,7 +9,6 @@
 #define NOOP 0
 #define LEFT 1
 #define RIGHT 2
-#define MAX_BALL_SPEED 448
 #define HALF_PADDLE_WIDTH 31
 #define Y_OFFSET 50
 #define TICK_RATE 1.0f/60.0f
@@ -55,9 +54,13 @@ typedef struct Breakout {
     float* brick_y;
     float* brick_states;
     int balls_fired;
+    float initial_paddle_width;
     float paddle_width;
     float paddle_height;
+    float paddle_speed;
     float ball_speed;
+    float initial_ball_speed;
+    float max_ball_speed;
     int hits;
     int width;
     int height;
@@ -289,15 +292,13 @@ bool calc_paddle_ball_collisions(Breakout* env, CollisionInfo* collision_info) {
     collision_info->brick_index = BRICK_INDEX_PADDLE_COLLISION;
 
     env->hit_brick = false;
-    float relative_intersection = ((env->ball_x +
-                                    env->ball_width / 2) -
-                                   env->paddle_x) /
-                                  env->paddle_width;
+    float relative_intersection = (
+        (env->ball_x + env->ball_width / 2) - env->paddle_x) / env->paddle_width;
     float angle = -base_angle + relative_intersection * 2 * base_angle;
     env->ball_vx = sin(angle) * env->ball_speed * TICK_RATE;
     env->ball_vy = -cos(angle) * env->ball_speed * TICK_RATE;
     env->hits += 1;
-    if (env->hits % 4 == 0 && env->ball_speed < MAX_BALL_SPEED) {
+    if (env->hits % 4 == 0 && env->ball_speed < env->max_ball_speed) {
         env->ball_speed += 64;
     }
     if (env->score == env->half_max_score) {
@@ -336,12 +337,16 @@ void calc_all_wall_collisions(Breakout* env, CollisionInfo* collision_info) {
 // With rare floating point conditions, the ball could escape the bounds.
 // Let's handle that explicitly.
 void check_wall_bounds(Breakout* env) {
-    if (env->ball_x < 0)
-        env->ball_x += MAX_BALL_SPEED * 1.1f * TICK_RATE;
-    if (env->ball_x > env->width)
-        env->ball_x -= MAX_BALL_SPEED * 1.1f * TICK_RATE;
-    if (env->ball_y < 0)
-        env->ball_y += MAX_BALL_SPEED * 1.1f * TICK_RATE;
+    float offset = env->max_ball_speed * 1.1f * TICK_RATE;
+    if (env->ball_x < 0) {
+        env->ball_x += offset;
+    }
+    if (env->ball_x > env->width) {
+        env->ball_x -= offset;
+    }
+    if (env->ball_y < 0) {
+        env->ball_y += offset;
+    }
 }
 
 void destroy_brick(Breakout* env, int brick_idx) {
@@ -353,7 +358,7 @@ void destroy_brick(Breakout* env, int brick_idx) {
     env->rewards[0] += gained_points;
 
     if (brick_idx / env->brick_cols < 3) {
-        env->ball_speed = MAX_BALL_SPEED;
+        env->ball_speed = env->max_ball_speed;
     }
 }
 
@@ -393,8 +398,8 @@ void reset_round(Breakout* env) {
     env->balls_fired = 0;
     env->hit_brick = false;
     env->hits = 0;
-    env->ball_speed = 256;
-    env->paddle_width = 2 * HALF_PADDLE_WIDTH;
+    env->ball_speed = env->initial_ball_speed;
+    env->paddle_width = env->initial_paddle_width;
 
     env->paddle_x = env->width / 2.0 - env->paddle_width / 2;
     env->paddle_y = env->height - env->paddle_height - 10;
@@ -437,7 +442,7 @@ void step_frame(Breakout* env, float action) {
     if (env->continuous){
         act = action;
     }
-    env->paddle_x += act * 620 * TICK_RATE;
+    env->paddle_x += act * env->paddle_speed * TICK_RATE;
     if (env->paddle_x <= 0){
         env->paddle_x = fmaxf(0, env->paddle_x);
     } else {

pufferlib/ocean/breakout/breakout.py

@@ -1,10 +1,3 @@
-'''High-perf Pong
-
-Inspired from https://gist.github.com/Yttrmin/18ecc3d2d68b407b4be1
-& https://jair.org/index.php/jair/article/view/10819/25823
-& https://www.youtube.com/watch?v=PSQt5KGv7Vk
-'''
-
 import numpy as np
 import gymnasium
 
@@ -17,7 +10,10 @@ def __init__(self, num_envs=1, render_mode=None,
             paddle_width=62, paddle_height=8,
             ball_width=32, ball_height=32,
             brick_width=32, brick_height=12,
-            brick_rows=6, brick_cols=18, continuous=False, log_interval=128,
+            brick_rows=6, brick_cols=18,
+            initial_ball_speed=256, max_ball_speed=448,
+            paddle_speed=620,
+            continuous=False, log_interval=128,
             buf=None, seed=0):
         self.single_observation_space = gymnasium.spaces.Box(low=0, high=1,
             shape=(10 + brick_rows*brick_cols,), dtype=np.float32)
@@ -40,10 +36,14 @@ def __init__(self, num_envs=1, render_mode=None,
             self.actions = self.actions.astype(np.float32)
 
         self.c_envs = binding.vec_init(self.observations, self.actions, self.rewards,
-            self.terminals, self.truncations, num_envs, seed, frameskip=frameskip, width=width, height=height,
-            paddle_width=paddle_width, paddle_height=paddle_height, ball_width=ball_width, ball_height=ball_height,
-            brick_width=brick_width, brick_height=brick_height, brick_rows=brick_rows,
-            brick_cols=brick_cols, continuous=continuous
+            self.terminals, self.truncations, num_envs, seed, frameskip=frameskip,
+            width=width, height=height, paddle_width=paddle_width,
+            paddle_height=paddle_height, ball_width=ball_width,
+            ball_height=ball_height, brick_width=brick_width,
+            brick_height=brick_height, brick_rows=brick_rows,
+            brick_cols=brick_cols, initial_ball_speed=initial_ball_speed,
+            max_ball_speed=max_ball_speed, paddle_speed=paddle_speed,
+            continuous=continuous
         )
 
     def reset(self, seed=0):