Fix d4rl conversion

d3rlpy/datasets.py

@@ -230,23 +230,61 @@ def get_d4rl(
         env = gym.make(env_name)
         dataset = env.get_dataset()
 
-        observations = dataset["observations"][1:]
-        actions = dataset["actions"][1:]
-        rewards = dataset["rewards"][:-1]
-        terminals = np.logical_and(
-            dataset["terminals"][:-1], np.logical_not(dataset["timeouts"][:-1])
-        )
-        episode_terminals = np.logical_or(
-            dataset["terminals"][:-1], dataset["timeouts"][:-1]
-        )
-        episode_terminals[-1] = 1.0
+        observations = []
+        actions = []
+        rewards = []
+        terminals = []
+        episode_terminals = []
+        episode_step = 0
+        cursor = 0
+        dataset_size = dataset["observations"].shape[0]
+        while cursor < dataset_size:
+            # collect data for step=t
+            observation = dataset["observations"][cursor]
+            action = dataset["actions"][cursor]
+            if episode_step == 0:
+                reward = 0.0
+            else:
+                reward = dataset["rewards"][cursor - 1]
+
+            observations.append(observation)
+            actions.append(action)
+            rewards.append(reward)
+            terminals.append(0.0)
+
+            episode_step += 1
+
+            if dataset["timeouts"][cursor]:
+                # skip adding the last step
+                episode_terminals.append(1.0)
+                episode_step = 0
+                cursor += 1
+                continue
+            else:
+                episode_terminals.append(0.0)
+
+            if dataset["terminals"][cursor]:
+                # collect data for step=t+1
+                dummy_observation = observation.copy()
+                dummy_action = action.copy()
+                next_reward = dataset["rewards"][cursor]
+
+                # the last observation is rarely used
+                observations.append(dummy_observation)
+                actions.append(dummy_action)
+                rewards.append(next_reward)
+                terminals.append(1.0)
+                episode_terminals.append(1.0)
+                episode_step = 0
+
+            cursor += 1
 
         mdp_dataset = MDPDataset(
-            observations=observations,
-            actions=actions,
-            rewards=rewards,
-            terminals=terminals,
-            episode_terminals=episode_terminals,
+            observations=np.array(observations, dtype=np.float32),
+            actions=np.array(actions, dtype=np.float32),
+            rewards=np.array(rewards, dtype=np.float32),
+            terminals=np.array(terminals, dtype=np.float32),
+            episode_terminals=np.array(episode_terminals, dtype=np.float32),
             create_mask=create_mask,
             mask_size=mask_size,
         )