Add aquarium env

Author: corentinlger

simulationsandbox/environments/aquarium.py

@@ -0,0 +1,154 @@
+from functools import partial
+
+import jax.numpy as jnp
+import matplotlib.pyplot as plt
+
+from jax import random, jit, vmap
+from flax import struct
+from matplotlib import colormaps
+
+from simulationsandbox.environments.base_env import BaseEnv, BaseEnvState
+
+N_DIMS = 3
+FISH_SPEED = 3.
+MOVE_SCALE = 1.
+FOOD_SPEED = 1.
+
+
+@struct.dataclass
+class Agents:
+    pos: jnp.array
+    velocity: jnp.array
+    alive: jnp.array
+    color: jnp.array
+    obs: jnp.array
+
+
+@struct.dataclass
+class Objects:
+    pos: jnp.array
+    velocity: jnp.array
+
+
+@struct.dataclass
+class AquiariumState(BaseEnvState):
+    time: int
+    grid_size: int
+    agents: Agents
+    objects: Objects
+
+
+def normal(theta):
+    return jnp.array([jnp.cos(theta), jnp.sin(theta)])
+
+normal = jit(vmap(normal))
+
+
+# Change the angle and speed of the agent a bit 
+def move(obs, key):
+    return random.normal(key, shape=(3,)) * MOVE_SCALE
+    return random.uniform(key, shape=(N_DIMS,), minval=-1, maxval=1) * MOVE_SCALE
+
+move = jit(vmap(move, in_axes=(0, 0)))
+
+
+class Aquarium(BaseEnv):
+    """ Minimalistic aquarium environmnent"""
+    def __init__(self, max_agents=10, max_objects=20, grid_size=20):
+        self.max_agents = max_agents
+        self.max_objects = max_objects
+        self.grid_size = grid_size
+
+    def init_state(self, num_agents, num_obs, key):
+        agents_key_pos, agents_key_vel, agents_color_key, objects_key_pos = random.split(key, 4)
+        fish_velocity = random.uniform(agents_key_vel, shape=(self.max_agents, N_DIMS), minval=-1, maxval=1)
+        fish_velocity = (fish_velocity / jnp.linalg.norm(fish_velocity)) * FISH_SPEED
+        # fish_velocity = fish_velocity  * FISH_SPEED
+
+        fish = Agents(
+            pos=random.uniform(key=agents_key_pos, shape=(self.max_agents, N_DIMS), minval=0, maxval=self.grid_size),
+            velocity=fish_velocity,
+            alive=jnp.hstack((jnp.ones(num_agents), jnp.zeros(self.max_agents - num_agents))),
+            color=random.uniform(key=agents_color_key, shape=(self.max_agents, 3), minval=0., maxval=1.),
+            obs=jnp.zeros((self.max_agents, num_obs))
+        )
+        
+        # Add food at the surface of the aquarium
+        x_y_food_pos=random.uniform(key=objects_key_pos, shape=(self.max_objects, 2), minval=0, maxval=self.grid_size) 
+        z_food_pos = jnp.full((self.max_objects, 1), fill_value=self.grid_size)
+        food_pos = jnp.concatenate((x_y_food_pos, z_food_pos), axis=1)
+        
+        food = Objects(
+            pos=food_pos,
+            velocity=jnp.tile(jnp.array([0., 0., -1]), (self.max_objects, 1)) * FOOD_SPEED,
+        )
+
+        aquarium_env =  AquiariumState(
+            time=0,
+            grid_size=self.grid_size,
+            agents=fish,
+            objects=food
+        )
+        
+        return aquarium_env
+    
+    @partial(jit, static_argnums=(0,))
+    def step(self, state, key):
+        keys = random.split(key, self.max_agents)
+        d_vel = move(state.agents.obs, keys)
+        velocity = state.agents.velocity + d_vel
+        velocity = (velocity / jnp.linalg.norm(velocity)) * FISH_SPEED
+        agents_pos = state.agents.pos + velocity
+
+        # Collide with walls
+        agents_pos = jnp.clip(agents_pos, 0, self.grid_size - 1)
+
+        # Update new state
+        time = state.time + 1
+        agents = state.agents.replace(pos=agents_pos, velocity=velocity)
+        state = state.replace(time=time, agents=agents)
+        return state
+    
+    def add_agent(self, state, agent_idx):
+        agents = state.agents.replace(alive=state.agents.alive.at[agent_idx].set(1.0))
+        state = state.replace(agents=agents)
+        return state
+    
+    def remove_agent(self, state, agent_idx):
+        agents = state.agents.replace(alive=state.agents.alive.at[agent_idx].set(0.0))
+        state = state.replace(agents=agents)
+        return state
+
+    @staticmethod
+    def render(state):
+        if not plt.fignum_exists(1):
+            plt.ion()
+            fig = plt.figure(figsize=(10, 10))
+            ax = fig.add_subplot(111, projection='3d')
+
+        plt.clf()
+
+        ax = plt.axes(projection='3d')
+
+        alive_agents = jnp.where(state.agents.alive != 0.0)
+        agents_x_pos = state.agents.pos[:, 0][alive_agents]
+        agents_y_pos = state.agents.pos[:, 1][alive_agents]
+        agents_z_pos = state.agents.pos[:, 2][alive_agents]
+        agents_colors = state.agents.color[alive_agents]
+
+        # TODO : see how to add cmap=colormaps["gist_rainbow"]
+        ax.scatter(agents_x_pos, agents_y_pos, agents_z_pos, c=agents_colors, marker="o", label="Fish")
+
+        ax.set_title("Multi-Agent Simulation")
+        ax.set_xlabel("X-axis")
+        ax.set_ylabel("Y-axis")
+        ax.set_zlabel("Z-axis")
+
+        ax.set_xlim(0, state.grid_size)
+        ax.set_ylim(0, state.grid_size)
+        ax.set_zlim(0, state.grid_size)
+
+        ax.legend()
+
+        plt.draw()
+        plt.pause(0.001)

simulationsandbox/utils/envs.py

@@ -1,8 +1,10 @@
 from simulationsandbox.environments.two_d_example_env import TwoDEnv
 from simulationsandbox.environments.three_d_example_env import ThreeDEnv
 from simulationsandbox.environments.lake_env import LakeEnv
+from simulationsandbox.environments.aquarium import Aquarium
 
 ENVS = {"two_d": TwoDEnv,
         "three_d": ThreeDEnv,
-        "lake": LakeEnv
+        "lake": LakeEnv,
+        "aquarium": Aquarium
         }