multiprocess verify

scenarionet/tests/local_test/_test_combine_dataset_local.py

@@ -13,7 +13,7 @@ def _test_combine_dataset():
 
     combine_path = os.path.join(SCENARIONET_DATASET_PATH, "combined_dataset")
     combine_multiple_dataset(combine_path, *dataset_paths, force_overwrite=True, try_generate_missing_file=True)
-    success, result = verify_loading_into_metadrive(combine_path, steps_to_run=250)
+    success, result = verify_loading_into_metadrive(combine_path, "./", steps_to_run=250)
     assert success
 
 

scenarionet/verifier/utils.py

@@ -1,5 +1,6 @@
 import json
 import logging
+import multiprocessing
 import os
 
 logger = logging.getLogger(__name__)
@@ -7,20 +8,46 @@ import tqdm
 from metadrive.envs.scenario_env import ScenarioEnv
 from metadrive.policy.replay_policy import ReplayEgoCarPolicy
 from metadrive.scenario.utils import get_number_of_scenarios
+from functools import partial
 
 
-def verify_loading_into_metadrive(dataset_path, result_save_dir=None, steps_to_run=0):
+def verify_loading_into_metadrive(dataset_path, result_save_dir, steps_to_run=300, num_workers=8):
-    print("================ Begin Scenario Loading Verification for {} ================ \n".format(result_save_dir))
-    scenario_num = get_number_of_scenarios(dataset_path)
     if result_save_dir is not None:
         assert os.path.exists(result_save_dir) and os.path.isdir(
             result_save_dir), "Argument result_save_dir must be an existing dir"
+    num_scenario = get_number_of_scenarios(dataset_path)
+    argument_list = []
+
+    func = partial(_loading_into_metadrive,
+                   dataset_path=dataset_path,
+                   result_save_dir=result_save_dir,
+                   steps_to_run=steps_to_run)
+
+    num_scenario_each_worker = int(num_scenario / num_workers)
+    for i in range(num_workers):
+        if i == num_workers - 1:
+            num_scenario_each_worker = num_scenario - num_scenario_each_worker * (num_workers - 1)
+        argument_list.append([i * num_scenario_each_worker, num_scenario_each_worker])
+
+    with multiprocessing.Pool(num_workers) as p:
+        result = list(p.imap(func, argument_list))
+    if all([i[0] for i in result]):
+        print("All scenarios can be loaded successfully!")
+    else:
+        print("Fail to load all scenarios, see log for more details! Number of logs: {}".format(num_workers))
+
+
+def _loading_into_metadrive(start_scenario_index, num_scenario, dataset_path, result_save_dir, steps_to_run):
+    print("================ Begin Scenario Loading Verification for {} ================ \n".format(result_save_dir))
+    assert os.path.exists(result_save_dir) and os.path.isdir(
+        result_save_dir), "Argument result_save_dir must be an existing dir"
     success = True
     env = ScenarioEnv(
         {
             "agent_policy": ReplayEgoCarPolicy,
-            "num_scenarios": scenario_num,
+            "num_scenarios": num_scenario,
             "horizon": 1000,
+            "start_scenario_index": start_scenario_index,
             "no_static_vehicles": False,
             "data_directory": dataset_path,
         }
@@ -28,10 +55,13 @@ def verify_loading_into_metadrive(dataset_path, result_save_dir=None, steps_to_r
     logging.disable(logging.WARNING)
     error_files = []
     try:
-        for i in tqdm.tqdm(range(scenario_num)):
+        for i in tqdm.tqdm(range(num_scenario),
+                           desc="Scenarios: {}-{}".format(start_scenario_index, start_scenario_index + num_scenario)):
             env.reset(force_seed=i)
             for i in range(steps_to_run):
-                env.step([0, 0])
+                o, r, d, i = env.step([0, 0])
+                if d:
+                    assert i["arrive_dest"], "Can not arrive destination"
     except Exception as e:
         file_name = env.engine.data_manager.summary_lookup[i]
         file_path = os.path.join(dataset_path, env.engine.data_manager.mapping[file_name], file_name)
@@ -42,7 +72,9 @@ def verify_loading_into_metadrive(dataset_path, result_save_dir=None, steps_to_r
     finally:
         env.close()
     if result_save_dir is not None:
-        with open(os.path.join(result_save_dir, "error_scenarios_{}.json".format(os.path.basename(dataset_path))),
+        file_name = "error_scenarios_{}_{}_{}.json".format(os.path.basename(dataset_path),
-                  "w+") as f:
+                                                           start_scenario_index,
+                                                           start_scenario_index + num_scenario)
+        with open(os.path.join(result_save_dir, file_name), "w+") as f:
             json.dump(error_files, f)
     return success, error_files