Tune CLI

python/ray/tune/logger.py

@@ -17,15 +17,8 @@
 
 logger = logging.getLogger(__name__)
 
-try:
-    import tensorflow as tf
-    use_tf150_api = (distutils.version.LooseVersion(tf.VERSION) >=
-                     distutils.version.LooseVersion("1.5.0"))
-except ImportError:
-    tf = None
-    use_tf150_api = True
-    logger.warning("Couldn't import TensorFlow - "
-                   "disabling TensorBoard logging.")
+tf = None
+use_tf150_api = True
 
 
 class Logger(object):
@@ -190,6 +183,15 @@ def to_tf_values(result, path):
 
 class _TFLogger(Logger):
     def _init(self):
+        try:
+            global tf, use_tf150_api
+            import tensorflow
+            tf = tensorflow
+            use_tf150_api = (distutils.version.LooseVersion(tf.VERSION) >=
+                             distutils.version.LooseVersion("1.5.0"))
+        except ImportError:
+            logger.warning("Couldn't import TensorFlow - "
+                           "disabling TensorBoard logging.")
         self._file_writer = tf.summary.FileWriter(self.logdir)
 
     def on_result(self, result):

python/ray/tune/scripts.py

@@ -8,10 +8,9 @@
 import os
 from datetime import datetime
 
-
 import pandas as pd
-
 from ray.tune.trial import Trial
+from tabulate import tabulate
 
 
 def _flatten_dict(dt):
@@ -34,13 +33,8 @@ def cli():
     pass
 
 
-DEFAULT_EXPERIMENT_INFO_KEYS = (
-    "trial_name",
-    "trial_id",
-    "status",
-    "num_failures",
-    "logdir"
-)
+DEFAULT_EXPERIMENT_INFO_KEYS = ("trial_name", "trial_id", "status",
+                                "num_failures", "logdir")
 
 DEFAULT_PROJECT_INFO_KEYS = (
     "name",
@@ -52,38 +46,47 @@ def cli():
 )
 
 
-def _list_trials(experiment_path, info_keys=DEFAULT_EXPERIMENT_INFO_KEYS):
+def _list_trials(experiment_path, sort,
+                 info_keys=DEFAULT_EXPERIMENT_INFO_KEYS):
     experiment_path = os.path.expanduser(experiment_path)
     globs = glob.glob(os.path.join(experiment_path, "experiment_state*.json"))
     filename = max(list(globs))
 
     with open(filename) as f:
         experiment_state = json.load(f)
 
-    checkpoints = pd.DataFrame.from_records(experiment_state['checkpoints'])
+    checkpoints_df = pd.DataFrame(
+        experiment_state["checkpoints"])[list(info_keys)]
+    if "logdir" in checkpoints_df.columns:
+        checkpoints_df["logdir"] = checkpoints_df["logdir"].str.replace(
+            experiment_path, '')
+    if sort:
+        checkpoints_df = checkpoints_df.sort_values(by=sort)
+    print(tabulate(checkpoints_df, headers="keys", tablefmt="psql"))
+
     # TODO(hartikainen): The logdir is often too verbose to be viewed in a
     # table.
-    checkpoints['logdir'] = checkpoints['logdir'].str.replace(
-        experiment_path, '')
-
-    print(checkpoints[list(info_keys)].to_string())
+    # checkpoints = pd.DataFrame.from_records(experiment_state['checkpoints'])
+    # checkpoints['logdir'] = checkpoints['logdir'].str.replace(
+    #     experiment_path, '')
+    # print(checkpoints[list(info_keys)].to_string())
 
 
 @cli.command()
 @click.argument("experiment_path", required=True, type=str)
-def list_trials(experiment_path):
-    _list_trials(experiment_path)
+@click.option(
+    '--sort', default=None, type=str, help='Select which column to sort on.')
+def list_trials(experiment_path, sort):
+    _list_trials(experiment_path, sort)
 
 
-def _list_experiments(project_path, info_keys=DEFAULT_PROJECT_INFO_KEYS):
+def _list_experiments(project_path, sort, info_keys=DEFAULT_PROJECT_INFO_KEYS):
     base, experiment_paths, _ = list(os.walk(project_path))[0]  # clean this
 
     experiment_data_collection = []
     for experiment_path in experiment_paths:
-        experiment_state_path = glob.glob(os.path.join(
-            base,
-            experiment_path,
-            "experiment_state*.json"))
+        experiment_state_path = glob.glob(
+            os.path.join(base, experiment_path, "experiment_state*.json"))
 
         if not experiment_state_path:
             # TODO(hartikainen): Print some warning?
@@ -94,29 +97,34 @@ def _list_experiments(project_path, info_keys=DEFAULT_PROJECT_INFO_KEYS):
 
         checkpoints = pd.DataFrame(experiment_state["checkpoints"])
         runner_data = experiment_state["runner_data"]
-        timestamp = experiment_state["timestamp"]
+        timestamp = experiment_state.get("timestamp")
 
         experiment_data = {
             "name": experiment_path,
-            "start_time": runner_data["_start_time"],
-            "timestamp": datetime.fromtimestamp(timestamp),
+            "start_time": runner_data.get("_start_time"),
+            "timestamp": datetime.fromtimestamp(timestamp)
+            if timestamp else None,
             "total_trials": checkpoints.shape[0],
             "running_trials": (checkpoints["status"] == Trial.RUNNING).sum(),
             "terminated_trials": (
                 checkpoints["status"] == Trial.TERMINATED).sum(),
             "error_trials": (checkpoints["status"] == Trial.ERROR).sum(),
-         }
+        }
 
         experiment_data_collection.append(experiment_data)
 
-    info_dataframe = pd.DataFrame(experiment_data_collection)
-    print(info_dataframe[list(info_keys)].to_string())
+    info_df = pd.DataFrame(experiment_data_collection)
+    if sort:
+        info_df = info_df.sort_values(by=sort)
+    print(tabulate(info_df, headers="keys", tablefmt="psql"))
 
 
 @cli.command()
 @click.argument("project_path", required=True, type=str)
-def list_experiments(project_path):
-    _list_experiments(project_path)
+@click.option(
+    '--sort', default=None, type=str, help='Select which column to sort on.')
+def list_experiments(project_path, sort):
+    _list_experiments(project_path, sort)
 
 
 cli.add_command(list_trials, name="ls")

python/ray/tune/suggest/bayesopt.py

@@ -4,13 +4,16 @@
 
 import copy
 
-try:
-    import bayes_opt as byo
-except Exception:
-    byo = None
-
 from ray.tune.suggest.suggestion import SuggestionAlgorithm
 
+byo = None
+
+
+def _import_bayesopt():
+    global byo
+    import bayes_opt
+    byo = bayes_opt
+
 
 class BayesOptSearch(SuggestionAlgorithm):
     """A wrapper around BayesOpt to provide trial suggestions.

python/ray/tune/suggest/hyperopt.py

@@ -6,17 +6,19 @@
 import copy
 import logging
 
-try:
-    hyperopt_logger = logging.getLogger("hyperopt")
-    hyperopt_logger.setLevel(logging.WARNING)
-    import hyperopt as hpo
-    from hyperopt.fmin import generate_trials_to_calculate
-except Exception:
-    hpo = None
-
 from ray.tune.error import TuneError
 from ray.tune.suggest.suggestion import SuggestionAlgorithm
 
+hpo = None
+
+
+def _import_hyperopt():
+    global hpo
+    hyperopt_logger = logging.getLogger("hyperopt")
+    hyperopt_logger.setLevel(logging.WARNING)
+    import hyperopt
+    hpo = hyperopt
+
 
 class HyperOptSearch(SuggestionAlgorithm):
     """A wrapper around HyperOpt to provide trial suggestions.
@@ -73,6 +75,7 @@ def __init__(self,
                  reward_attr="episode_reward_mean",
                  points_to_evaluate=None,
                  **kwargs):
+        _import_hyperopt()
         assert hpo is not None, "HyperOpt must be installed!"
         assert type(max_concurrent) is int and max_concurrent > 0
         self._max_concurrent = max_concurrent
@@ -84,7 +87,7 @@ def __init__(self,
             self._points_to_evaluate = 0
         else:
             assert type(points_to_evaluate) == list
-            self._hpopt_trials = generate_trials_to_calculate(
+            self._hpopt_trials = hpo.fmin.generate_trials_to_calculate(
                 points_to_evaluate)
             self._hpopt_trials.refresh()
             self._points_to_evaluate = len(points_to_evaluate)

python/ray/tune/trial_runner.py

@@ -547,8 +547,12 @@ def __getstate__(self):
         """
         state = self.__dict__.copy()
         for k in [
-                "_trials", "_stop_queue", "_server", "_search_alg",
-                "_scheduler_alg", "trial_executor",
+                "_trials",
+                "_stop_queue",
+                "_server",
+                "_search_alg",
+                "_scheduler_alg",
+                "trial_executor",
         ]:
             del state[k]
         state["_start_time"] = self._start_time.timestamp()
@@ -561,7 +565,8 @@ def __setstate__(self, state):
         session = state.pop("_session")
         self.__dict__.setdefault("_session", session)
         start_time = state.pop("_start_time")
-        self.__dict__.setdefault("_start_time", datetime.fromtimestamp(start_time))
+        self.__dict__.setdefault("_start_time",
+                                 datetime.fromtimestamp(start_time))
 
         self.__dict__.update(state)
         if launch_web_server:

python/setup.py

@@ -169,8 +169,7 @@ def find_version(*filepath):
     entry_points={
         "console_scripts": [
             "ray=ray.scripts.scripts:main",
-            "rllib=ray.rllib.scripts:cli [rllib]",
-            "tune=ray.tune.scripts:cli"
+            "rllib=ray.rllib.scripts:cli [rllib]", "tune=ray.tune.scripts:cli"
         ]
     },
     include_package_data=True,