add docs for actions/observations,

refactor basic usage and rename stuff

docs/src/usage/actions.rst

@@ -3,4 +3,44 @@
 Actions
 =====================
 
-Work in progress on actions.
+Several **types of actions** for longitudinal and lateral control are available in every track.
+
+Lateral actions:
+
+- `steering_angle`: the action sets the target steering angle of the vehicle in `rad`, which is then converted to steering speed.
+
+- `steering_speed`: the action directly sets the steering speed of the vehicle in `rad/s`.
+
+Longitudinal actions:
+
+- `speed`: the action sets the target speed of the vehicle in `m/s`, which is then converted to acceleration.
+
+- `accl`: the action directly sets the vehicle acceleration in `m/s^2`.
+
+**Note:** All the agents will have the same observation type.
+
+Actions Configuration
+---------------------
+The environment comes with a *default* action type, which can be changed using the environment configuration.
+
+Actions can be configured at the environment creation:
+
+.. code:: python
+
+    import gymnasium as gym
+
+    env = gym.make(
+		"f110_gym:f110-v0",
+		config={
+			"control_input": ["accl", "steering_speed"]
+		})
+    obs, infos = env.reset()
+or after the environment creation:
+
+.. code:: python
+
+    import gymnasium as gym
+
+    env = gym.make("f110_gym:f110-v0")
+	env.configure({"control_input": ["speed", "steering_angle"]})
+    obs, infos = env.reset()

docs/src/usage/basic_usage.rst

@@ -1,6 +1,6 @@
 .. _basic_usage:
 
-Basic Usage Example
+Basic example
 =====================
 
 The environment can work out of the box without too much customization.

docs/src/usage/index.rst

@@ -1,13 +1,44 @@
-Usage
+Getting Started
 ==============
 
+Making an environment
+----------------------
+
+This is a quick example of how to make a simple environment:
+
+.. code:: python
+
+    import gymnasium as gym
+
+    env = gym.make("f110_gym:f110-v0", render_mode="human")
+    obs, infos = env.reset()
+
+    while not done:
+        # sample random action
+        actions = env.action_space.sample()
+        # step simulation
+        obs, step_reward, done, info = racecar_env.step(actions)
+        env.render()
+
+Configuration
+-------------
+
+Here is the list of configurable options for the environment:
+- Track
+- Observations
+- Actions
+- Rewards
+- Dynamics
+- Rendering
+- Extra tools
+
 .. toctree::
-  :caption: Usage
-  :maxdepth: 2
+  :caption: Getting started
+  :maxdepth: 1
 
   basic_usage
-  actions
   observations
+  actions
   rewards
   dynamics
   customized_usage

docs/src/usage/observations.rst

@@ -3,4 +3,45 @@
 Observations
 =====================
 
-Work in progress.
+For the environment, several **types of observations** can be used.
+
+- `original` : Original observation from the old simulator. This is default for compatibility.
+- `kinematic_state` : Kinematic state observation, which includes `pose_x, pose_y, delta, linear_vel_x, pose_theta`.
+- `dynamic_state` : Dynamic state observation, which includes `pose_x, pose_y, delta, linear_vel_x, pose_theta, ang_vel_z, beta`.
+- `features` : Customisable observation, which includes all the features defined in the `features` argument.
+
+**Note:** All the agents will have the same observation type.
+
+Observations Configuration
+--------------------------
+Each environment comes with a *default* observation,
+which can be changed or customised using environment configurations.
+
+Observations can be configured at the environment creation:
+
+.. code:: python
+
+    import gymnasium as gym
+
+    env = gym.make(
+		"f110_gym:f110-v0",
+		config={
+			"observation_config": {
+				"type": "features",
+				"features": ["linear_vel_x", "scan"]
+			},
+		})
+    obs, infos = env.reset()
+or after the environment creation:
+
+.. code:: python
+
+    import gymnasium as gym
+
+    env = gym.make("f110_gym:f110-v0")
+    env.configure({
+		"observation_config": {
+			"type": "features",
+			"features": ["linear_vel_x", "scan"]}
+		})
+    obs, infos = env.reset()