[tune] Deprecate tune.function (#5601)

* remove tune function

* remove examples

* Update tune-usage.rst

doc/source/rllib-training.rst

@@ -286,11 +286,11 @@ You can provide callback functions to be called at points during policy evaluati
         config={
             "env": "CartPole-v0",
             "callbacks": {
-                "on_episode_start": tune.function(on_episode_start),
-                "on_episode_step": tune.function(on_episode_step),
-                "on_episode_end": tune.function(on_episode_end),
-                "on_train_result": tune.function(on_train_result),
-                "on_postprocess_traj": tune.function(on_postprocess_traj),
+                "on_episode_start": on_episode_start,
+                "on_episode_step": on_episode_step,
+                "on_episode_end": on_episode_end,
+                "on_train_result": on_train_result,
+                "on_postprocess_traj": on_postprocess_traj,
             },
         },
     )
@@ -377,7 +377,7 @@ Approach 2: Use the callbacks API to update the environment on new training resu
         config={
             "env": YourEnv,
             "callbacks": {
-                "on_train_result": tune.function(on_train_result),
+                "on_train_result": on_train_result,
             },
         },
     )

doc/source/tune-usage.rst

@@ -167,9 +167,8 @@ The following shows grid search over two nested parameters combined with random
         }
     )
 
-
-.. note::
-    Use ``tune.sample_from(...)`` to sample from a function during trial variant generation. If you need to pass a literal function in your config, use ``tune.function(...)`` to escape it.
+.. note::	
+    Use ``tune.sample_from(...)`` to sample from a function during trial variant generation.
 
 For more information on variant generation, see `basic_variant.py <https://github.com/ray-project/ray/blob/master/python/ray/tune/suggest/basic_variant.py>`__.
 
@@ -177,8 +176,7 @@ Custom Trial Names
 ------------------
 
 To specify custom trial names, you can pass use the ``trial_name_creator`` argument
-to `tune.run`.  This takes a function with the following signature, and
-be sure to wrap it with `tune.function`:
+to `tune.run`.  This takes a function with the following signature:
 
 .. code-block:: python
 
@@ -196,7 +194,7 @@ be sure to wrap it with `tune.function`:
         MyTrainableClass,
         name="example-experiment",
         num_samples=1,
-        trial_name_creator=tune.function(trial_name_string)
+        trial_name_creator=trial_name_string
     )
 
 An example can be found in `logging_example.py <https://github.com/ray-project/ray/blob/master/python/ray/tune/examples/logging_example.py>`__.
@@ -496,7 +494,7 @@ Uploading/Syncing
 Tune automatically syncs the trial folder on remote nodes back to the head node. This requires the ray cluster to be started with the `autoscaler <autoscaling.html>`__.
 By default, local syncing requires rsync to be installed. You can customize the sync command with the ``sync_to_driver`` argument in ``tune.run`` by providing either a function or a string.
 
-If a string is provided, then it must include replacement fields ``{source}`` and ``{target}``, like ``rsync -savz -e "ssh -i ssh_key.pem" {source} {target}``. Alternatively, a function can be provided with the following signature (and must be wrapped with ``tune.function``):
+If a string is provided, then it must include replacement fields ``{source}`` and ``{target}``, like ``rsync -savz -e "ssh -i ssh_key.pem" {source} {target}``. Alternatively, a function can be provided with the following signature:
 
 .. code-block:: python
 
@@ -510,7 +508,7 @@ If a string is provided, then it must include replacement fields ``{source}`` an
     tune.run(
         MyTrainableClass,
         name="experiment_name",
-        sync_to_driver=tune.function(custom_sync_func),
+        sync_to_driver=custom_sync_func,
     )
 
 When syncing results back to the driver, the source would be a path similar to ``ubuntu@192.0.0.1:/home/ubuntu/ray_results/trial1``, and the target would be a local path.
@@ -524,7 +522,7 @@ You can customize this to specify arbitrary storages with the ``sync_to_cloud``
     tune.run(
         MyTrainableClass,
         name="experiment_name",
-        sync_to_cloud=tune.function(custom_sync_func),
+        sync_to_cloud=custom_sync_func,
     )
 
 Tune Client API

python/ray/tune/examples/logging_example.py

@@ -64,7 +64,7 @@ if __name__ == "__main__":
         MyTrainableClass,
         name="hyperband_test",
         num_samples=5,
-        trial_name_creator=tune.function(trial_str_creator),
+        trial_name_creator=trial_str_creator,
         loggers=[TestLogger],
         stop={"training_iteration": 1 if args.smoke_test else 99999},
         config={

python/ray/tune/sample.py

@@ -11,9 +11,6 @@ logger = logging.getLogger(__name__)
 class sample_from(object):
     """Specify that tune should sample configuration values from this function.
 
-    The use of function arguments in tune configs must be disambiguated by
-    either wrapped the function in tune.sample_from() or tune.function().
-
     Arguments:
         func: An callable function to draw a sample from.
     """
@@ -28,27 +25,10 @@ class sample_from(object):
         return "tune.sample_from({})".format(repr(self.func))
 
 
-class function(object):
-    """Wraps `func` to make sure it is not expanded during resolution.
-
-    The use of function arguments in tune configs must be disambiguated by
-    either wrapped the function in tune.sample_from() or tune.function().
-
-    Arguments:
-        func: A function literal.
-    """
-
-    def __init__(self, func):
-        self.func = func
-
-    def __call__(self, *args, **kwargs):
-        return self.func(*args, **kwargs)
-
-    def __str__(self):
-        return "tune.function({})".format(str(self.func))
-
-    def __repr__(self):
-        return "tune.function({})".format(repr(self.func))
+def function(func):
+    logger.warn("DeprecationWarning: wrapping {} with tune.function() is no "
+                "longer needed".format(func))
+    return func
 
 
 def uniform(*args, **kwargs):

python/ray/tune/suggest/variant_generator.py

@@ -162,9 +162,7 @@ def _resolve_lambda_vars(spec, lambda_vars):
                 error = e
             except Exception:
                 raise ValueError(
-                    "Failed to evaluate expression: {}: {}".format(path, fn) +
-                    ". If you meant to pass this as a function literal, use "
-                    "tune.function() to escape it.")
+                    "Failed to evaluate expression: {}: {}".format(path, fn))
             else:
                 _assign_value(spec, path, value)
                 resolved[path] = value
@@ -207,16 +205,7 @@ def _is_resolved(v):
 
 
 def _try_resolve(v):
-    if isinstance(v, types.FunctionType):
-        raise DeprecationWarning(
-            "Function values are ambiguous in Tune "
-            "configuations. Either wrap the function with "
-            "`tune.function(func)` to specify a function literal, or "
-            "`tune.sample_from(func)` to tell Tune to "
-            "sample values from the function during variant generation: "
-            "{}".format(v))
-        return False, v
-    elif isinstance(v, sample_from):
+    if isinstance(v, sample_from):
         # Function to sample from
         return False, v.func
     elif isinstance(v, dict) and len(v) == 1 and "eval" in v:

python/ray/tune/syncer.py

@@ -15,7 +15,6 @@ try:  # py3
 except ImportError:  # py2
     from pipes import quote
 
-from ray.tune.sample import function as tune_function
 from ray.tune.error import TuneError
 from ray.tune.log_sync import log_sync_template, NodeSyncMixin
 
@@ -169,8 +168,7 @@ class CommandSyncer(BaseSyncer):
 def _get_sync_cls(sync_function):
     if not sync_function:
         return
-    if isinstance(sync_function, types.FunctionType) or isinstance(
-            sync_function, tune_function):
+    if isinstance(sync_function, types.FunctionType):
         return BaseSyncer
     elif isinstance(sync_function, str):
         return CommandSyncer

python/ray/tune/tests/test_trial_runner.py

@@ -972,8 +972,8 @@ class RunExperimentTest(unittest.TestCase):
                 "stop": {
                     "training_iteration": 1
                 },
-                "trial_name_creator": tune.function(
-                    lambda t: "{}_{}_321".format(t.trainable_name, t.trial_id))
+                "trial_name_creator":
+                    lambda t: "{}_{}_321".format(t.trainable_name, t.trial_id)
             }
         })
         self.assertEquals(
@@ -1113,7 +1113,7 @@ class TestSyncFunctionality(unittest.TestCase):
                 "training_iteration": 1
             },
             upload_dir=tmpdir2,
-            sync_to_cloud=tune.function(sync_func)).trials
+            sync_to_cloud=sync_func).trials
         test_file_path = glob.glob(os.path.join(tmpdir2, "foo", "*.json"))
         self.assertTrue(test_file_path)
         shutil.rmtree(tmpdir)
@@ -1134,7 +1134,7 @@ class TestSyncFunctionality(unittest.TestCase):
             stop={
                 "training_iteration": 1
             },
-            sync_to_driver=tune.function(sync_func_driver)).trials
+            sync_to_driver=sync_func_driver).trials
         test_file_path = os.path.join(trial.logdir, "test.log2")
         self.assertFalse(os.path.exists(test_file_path))
 
@@ -1147,7 +1147,7 @@ class TestSyncFunctionality(unittest.TestCase):
                 stop={
                     "training_iteration": 1
                 },
-                sync_to_driver=tune.function(sync_func_driver)).trials
+                sync_to_driver=sync_func_driver).trials
         test_file_path = os.path.join(trial.logdir, "test.log2")
         self.assertTrue(os.path.exists(test_file_path))
         os.remove(test_file_path)
@@ -1166,8 +1166,8 @@ class TestSyncFunctionality(unittest.TestCase):
                         "training_iteration": 1
                     },
                     "upload_dir": "test",
-                    "sync_to_driver": tune.function(sync_func),
-                    "sync_to_cloud": tune.function(sync_func)
+                    "sync_to_driver": sync_func,
+                    "sync_to_cloud": sync_func
                 }).trials
             self.assertEqual(mock_sync.call_count, 0)
 
@@ -2273,11 +2273,9 @@ class TrialRunnerTest(unittest.TestCase):
         ray.init()
         trial = Trial(
             "__fake",
-            config={
-                "callbacks": {
-                    "on_episode_start": tune.function(lambda i: i),
-                }
-            },
+            config={"callbacks": {
+                "on_episode_start": lambda i: i,
+            }},
             checkpoint_freq=1)
         tmpdir = tempfile.mkdtemp()
         runner = TrialRunner(local_checkpoint_dir=tmpdir, checkpoint_period=0)

python/ray/tune/tests/test_tune_server.py

@@ -8,7 +8,6 @@ import subprocess
 import json
 
 import ray
-from ray import tune
 from ray.rllib import _register_all
 from ray.tune.trial import Trial, Resources
 from ray.tune.web_server import TuneClient
@@ -94,11 +93,9 @@ class TuneServerSuite(unittest.TestCase):
             "__fake",
             trial_id="function_trial",
             stopping_criterion={"training_iteration": 3},
-            config={
-                "callbacks": {
-                    "on_episode_start": tune.function(lambda x: None)
-                }
-            })
+            config={"callbacks": {
+                "on_episode_start": lambda x: None
+            }})
         runner.add_trial(test_trial)
 
         for i in range(3):

python/ray/tune/trial_runner.py

@@ -11,6 +11,7 @@ import os
 import re
 import time
 import traceback
+import types
 
 import ray.cloudpickle as cloudpickle
 from ray.tune import TuneError
@@ -19,7 +20,6 @@ from ray.tune.result import (TIME_THIS_ITER_S, RESULT_DUPLICATE,
                              SHOULD_CHECKPOINT)
 from ray.tune.syncer import get_syncer
 from ray.tune.trial import Trial, Checkpoint
-from ray.tune.sample import function
 from ray.tune.schedulers import FIFOScheduler, TrialScheduler
 from ray.tune.suggest import BasicVariantGenerator
 from ray.tune.util import warn_if_slow, flatten_dict
@@ -48,7 +48,7 @@ def _find_newest_ckpt(ckpt_dir):
 
 class _TuneFunctionEncoder(json.JSONEncoder):
     def default(self, obj):
-        if isinstance(obj, function):
+        if isinstance(obj, types.FunctionType):
             return self._to_cloudpickle(obj)
         try:
             return super(_TuneFunctionEncoder, self).default(obj)

rllib/agents/dqn/dqn.py

@@ -4,7 +4,6 @@ from __future__ import print_function
 
 import logging
 
-from ray import tune
 from ray.rllib.agents.trainer import with_common_config
 from ray.rllib.agents.trainer_template import build_trainer
 from ray.rllib.agents.dqn.dqn_policy import DQNTFPolicy
@@ -174,8 +173,7 @@ def check_config_and_setup_param_noise(config):
             if start_callback:
                 start_callback(info)
 
-        config["callbacks"]["on_episode_start"] = tune.function(
-            on_episode_start)
+        config["callbacks"]["on_episode_start"] = on_episode_start
         if config["callbacks"]["on_episode_end"]:
             end_callback = config["callbacks"]["on_episode_end"]
         else:
@@ -191,7 +189,7 @@ def check_config_and_setup_param_noise(config):
             if end_callback:
                 end_callback(info)
 
-        config["callbacks"]["on_episode_end"] = tune.function(on_episode_end)
+        config["callbacks"]["on_episode_end"] = on_episode_end
 
 
 def get_initial_state(config):

rllib/evaluation/rollout_worker.py

@@ -259,10 +259,7 @@ class RolloutWorker(EvaluatorInterface):
         policy_mapping_fn = (policy_mapping_fn
                              or (lambda agent_id: DEFAULT_POLICY_ID))
         if not callable(policy_mapping_fn):
-            raise ValueError(
-                "Policy mapping function not callable. If you're using Tune, "
-                "make sure to escape the function with tune.function() "
-                "to prevent it from being evaluated as an expression.")
+            raise ValueError("Policy mapping function not callable?")
         self.env_creator = env_creator
         self.sample_batch_size = batch_steps * num_envs
         self.batch_mode = batch_mode

rllib/examples/centralized_critic.py

@@ -222,8 +222,7 @@ if __name__ == "__main__":
                     "pol2": (None, TwoStepGame.observation_space,
                              TwoStepGame.action_space, {}),
                 },
-                "policy_mapping_fn": tune.function(
-                    lambda x: "pol1" if x == 0 else "pol2"),
+                "policy_mapping_fn": lambda x: "pol1" if x == 0 else "pol2",
             },
             "model": {
                 "custom_model": "cc_model",

rllib/examples/centralized_critic_2.py

@@ -139,7 +139,7 @@ if __name__ == "__main__":
             "env": GlobalObsTwoStepGame,
             "batch_mode": "complete_episodes",
             "callbacks": {
-                "on_postprocess_traj": tune.function(fill_in_actions),
+                "on_postprocess_traj": fill_in_actions,
             },
             "num_workers": 0,
             "multiagent": {
@@ -149,8 +149,7 @@ if __name__ == "__main__":
                     "pol2": (None, GlobalObsTwoStepGame.observation_space,
                              GlobalObsTwoStepGame.action_space, {}),
                 },
-                "policy_mapping_fn": tune.function(
-                    lambda x: "pol1" if x == 0 else "pol2"),
+                "policy_mapping_fn": lambda x: "pol1" if x == 0 else "pol2",
             },
             "model": {
                 "custom_model": "cc_model",

rllib/examples/custom_metrics_and_callbacks.py

@@ -67,12 +67,12 @@ if __name__ == "__main__":
         config={
             "env": "CartPole-v0",
             "callbacks": {
-                "on_episode_start": tune.function(on_episode_start),
-                "on_episode_step": tune.function(on_episode_step),
-                "on_episode_end": tune.function(on_episode_end),
-                "on_sample_end": tune.function(on_sample_end),
-                "on_train_result": tune.function(on_train_result),
-                "on_postprocess_traj": tune.function(on_postprocess_traj),
+                "on_episode_start": on_episode_start,
+                "on_episode_step": on_episode_step,
+                "on_episode_end": on_episode_end,
+                "on_sample_end": on_sample_end,
+                "on_train_result": on_train_result,
+                "on_postprocess_traj": on_postprocess_traj,
             },
         },
         return_trials=True)

rllib/examples/multiagent_cartpole.py

@@ -108,7 +108,7 @@ if __name__ == "__main__":
             "num_sgd_iter": 10,
             "multiagent": {
                 "policies": policies,
-                "policy_mapping_fn": tune.function(
+                "policy_mapping_fn": (
                     lambda agent_id: random.choice(policy_ids)),
             },
         },

rllib/examples/multiagent_custom_policy.py

@@ -69,7 +69,7 @@ if __name__ == "__main__":
                     "pg_policy": (None, obs_space, act_space, {}),
                     "random": (RandomPolicy, obs_space, act_space, {}),
                 },
-                "policy_mapping_fn": tune.function(
+                "policy_mapping_fn": (
                     lambda agent_id: ["pg_policy", "random"][agent_id % 2]),
             },
         },

rllib/examples/rock_paper_scissors_multiagent.py

@@ -185,7 +185,7 @@ def run_heuristic_vs_learned(use_lstm=False, trainer="PG"):
                         }
                     }),
                 },
-                "policy_mapping_fn": tune.function(select_policy),
+                "policy_mapping_fn": select_policy,
             },
         })
 

rllib/examples/twostep_game.py

@@ -129,8 +129,7 @@ if __name__ == "__main__":
                                  "agent_id": 1,
                              }),
                 },
-                "policy_mapping_fn": tune.function(
-                    lambda x: "pol1" if x == 0 else "pol2"),
+                "policy_mapping_fn": lambda x: "pol1" if x == 0 else "pol2",
             },
         }
         group = False