#!/usr/bin/env python3
# Copyright (c) Meta Platforms, Inc. and affiliates.
#
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
import json
import logging
import warnings
from functools import partial
from logging import Logger
from typing import (
Any,
Callable,
Dict,
Iterable,
List,
Optional,
Sequence,
Set,
Tuple,
Type,
TypeVar,
Union,
)
import ax.service.utils.early_stopping as early_stopping_utils
import numpy as np
import pandas as pd
import torch
from ax.core.arm import Arm
from ax.core.base_trial import BaseTrial, TrialStatus
from ax.core.batch_trial import BatchTrial
from ax.core.data import Data
from ax.core.experiment import DataType, Experiment
from ax.core.generator_run import GeneratorRun
from ax.core.objective import MultiObjective, Objective
from ax.core.observation import ObservationFeatures
from ax.core.optimization_config import (
MultiObjectiveOptimizationConfig,
OptimizationConfig,
)
from ax.core.search_space import HierarchicalSearchSpace
from ax.core.trial import Trial
from ax.core.types import (
TEvaluationOutcome,
TModelPredictArm,
TParameterization,
TParamValue,
)
from ax.early_stopping.strategies import BaseEarlyStoppingStrategy
from ax.exceptions.constants import CHOLESKY_ERROR_ANNOTATION
from ax.exceptions.core import (
OptimizationComplete,
OptimizationShouldStop,
UnsupportedError,
UnsupportedPlotError,
)
from ax.exceptions.generation_strategy import MaxParallelismReachedException
from ax.global_stopping.strategies.base import BaseGlobalStoppingStrategy
from ax.modelbridge.dispatch_utils import choose_generation_strategy
from ax.modelbridge.generation_strategy import GenerationStrategy
from ax.modelbridge.modelbridge_utils import (
get_pending_observation_features_based_on_trial_status,
)
from ax.modelbridge.prediction_utils import predict_by_features
from ax.plot.base import AxPlotConfig
from ax.plot.contour import plot_contour
from ax.plot.feature_importances import plot_feature_importance_by_feature
from ax.plot.helper import _format_dict
from ax.plot.trace import optimization_trace_single_method
from ax.service.utils.best_point_mixin import BestPointMixin
from ax.service.utils.instantiation import InstantiationBase, ObjectiveProperties
from ax.service.utils.report_utils import exp_to_df
from ax.service.utils.with_db_settings_base import DBSettings, WithDBSettingsBase
from ax.storage.json_store.decoder import (
generation_strategy_from_json,
object_from_json,
)
from ax.storage.json_store.encoder import object_to_json
from ax.storage.json_store.registry import (
CORE_CLASS_DECODER_REGISTRY,
CORE_CLASS_ENCODER_REGISTRY,
CORE_DECODER_REGISTRY,
CORE_ENCODER_REGISTRY,
)
from ax.utils.common.docutils import copy_doc
from ax.utils.common.executils import retry_on_exception
from ax.utils.common.logger import _round_floats_for_logging, get_logger
from ax.utils.common.typeutils import checked_cast, checked_cast_complex, not_none
from botorch.utils.sampling import manual_seed
logger: Logger = get_logger(__name__)
AxClientSubclass = TypeVar("AxClientSubclass", bound="AxClient")
ROUND_FLOATS_IN_LOGS_TO_DECIMAL_PLACES: int = 6
# pyre-fixme[5]: Global expression must be annotated.
round_floats_for_logging = partial(
_round_floats_for_logging,
decimal_places=ROUND_FLOATS_IN_LOGS_TO_DECIMAL_PLACES,
)
[docs]class AxClient(WithDBSettingsBase, BestPointMixin, InstantiationBase):
"""
Convenience handler for management of experimentation cycle through a
service-like API. External system manages scheduling of the cycle and makes
calls to this client to get next suggestion in the experiment and log back
data from the evaluation of that suggestion.
Note: `AxClient` expects to only propose 1 arm (suggestion) per trial; support
for use cases that require use of batches is coming soon.
Two custom types used in this class for convenience are `TParamValue` and
`TParameterization`. Those are shortcuts for `Union[str, bool, float, int]`
and `Dict[str, Union[str, bool, float, int]]`, respectively.
Args:
generation_strategy: Optional generation strategy. If not set, one is
intelligently chosen based on properties of search space.
db_settings: Settings for saving and reloading the underlying experiment
to a database. Expected to be of type
ax.storage.sqa_store.structs.DBSettings and require SQLAlchemy.
enforce_sequential_optimization: Whether to enforce that when it is
reasonable to switch models during the optimization (as prescribed
by `num_trials` in generation strategy), Ax will wait for enough trials
to be completed with data to proceed. Defaults to True. If set to
False, Ax will keep generating new trials from the previous model
until enough data is gathered. Use this only if necessary;
otherwise, it is more resource-efficient to
optimize sequentially, by waiting until enough data is available to
use the next model.
random_seed: Optional integer random seed, set to fix the optimization
random seed for reproducibility. Works only for Sobol quasi-random
generator and for BoTorch-powered models. For the latter models, the
trials generated from the same optimization setup with the same seed,
will be mostly similar, but the exact parameter values may still vary
and trials latter in the optimizations will diverge more and more.
This is because a degree of randomness is essential for high performance
of the Bayesian optimization models and is not controlled by the seed.
Note: In multi-threaded environments, the random seed is thread-safe,
but does not actually guarantee reproducibility. Whether the outcomes
will be exactly the same for two same operations that use the random
seed, depends on whether the threads modify the random state in the
same order across the two operations.
torch_device: An optional `torch.device` object, used to choose the device
used for generating new points for trials. Works only for torch-based
models, such as GPEI. Ignored if a `generation_strategy` is passed in
manually. To specify the device for a custom `generation_strategy`,
pass in `torch_device` as part of `model_kwargs`. See
https://ax.dev/tutorials/generation_strategy.html for a tutorial on
generation strategies.
verbose_logging: Whether Ax should log significant optimization events,
defaults to `True`.
suppress_storage_errors: Whether to suppress SQL storage-related errors if
encounted. Only use if SQL storage is not important for the given use
case, since this will only log, but not raise, an exception if its
encountered while saving to DB or loading from it.
early_stopping_strategy: A ``BaseEarlyStoppingStrategy`` that determines
whether a trial should be stopped given the current state of
the experiment. Used in ``should_stop_trials_early``.
global_stopping_strategy: A ``BaseGlobalStoppingStrategy`` that determines
whether the full optimization should be stopped or not.
"""
BATCH_TRIAL_RAW_DATA_FORMAT_ERROR_MESSAGE = (
"Raw data must be a dict for batched trials."
)
TRIAL_RAW_DATA_FORMAT_ERROR_MESSAGE = (
"Raw data must be data for a single arm for non batched trials."
)
_experiment: Optional[Experiment] = None
def __init__(
self,
generation_strategy: Optional[GenerationStrategy] = None,
db_settings: Optional[DBSettings] = None,
enforce_sequential_optimization: bool = True,
random_seed: Optional[int] = None,
torch_device: Optional[torch.device] = None,
verbose_logging: bool = True,
suppress_storage_errors: bool = False,
early_stopping_strategy: Optional[BaseEarlyStoppingStrategy] = None,
global_stopping_strategy: Optional[BaseGlobalStoppingStrategy] = None,
) -> None:
super().__init__(
db_settings=db_settings,
suppress_all_errors=suppress_storage_errors,
)
if not verbose_logging:
logger.setLevel(logging.WARNING) # pragma: no cover
else:
logger.info(
"Starting optimization with verbose logging. To disable logging, "
"set the `verbose_logging` argument to `False`. Note that float "
"values in the logs are rounded to "
f"{ROUND_FLOATS_IN_LOGS_TO_DECIMAL_PLACES} decimal points."
)
if generation_strategy is not None and torch_device is not None:
warnings.warn(
"Both a `generation_strategy` and a `torch_device` were specified. "
"`torch_device` will be ignored. Instead, specify `torch_device` "
"by passing it in `model_kwargs` while creating the "
"`generation_strategy`.",
RuntimeWarning,
)
self._generation_strategy = generation_strategy
self._enforce_sequential_optimization = enforce_sequential_optimization
self._random_seed = random_seed
self._torch_device = torch_device
self._suppress_storage_errors = suppress_storage_errors
self._early_stopping_strategy = early_stopping_strategy
self._global_stopping_strategy = global_stopping_strategy
if random_seed is not None:
logger.warning(
f"Random seed set to {random_seed}. Note that this setting "
"only affects the Sobol quasi-random generator "
"and BoTorch-powered Bayesian optimization models. For the latter "
"models, setting random seed to the same number for two optimizations "
"will make the generated trials similar, but not exactly the same, "
"and over time the trials will diverge more."
)
# ------------------------ Public API methods. ------------------------
[docs] def create_experiment(
self,
parameters: List[
Dict[str, Union[TParamValue, Sequence[TParamValue], Dict[str, List[str]]]]
],
name: Optional[str] = None,
description: Optional[str] = None,
owners: Optional[List[str]] = None,
objective_name: Optional[str] = None,
minimize: Optional[bool] = None,
objectives: Optional[Dict[str, ObjectiveProperties]] = None,
parameter_constraints: Optional[List[str]] = None,
outcome_constraints: Optional[List[str]] = None,
status_quo: Optional[TParameterization] = None,
overwrite_existing_experiment: bool = False,
experiment_type: Optional[str] = None,
tracking_metric_names: Optional[List[str]] = None,
choose_generation_strategy_kwargs: Optional[Dict[str, Any]] = None,
support_intermediate_data: bool = False,
immutable_search_space_and_opt_config: bool = True,
is_test: bool = False,
metric_definitions: Optional[Dict[str, Dict[str, Any]]] = None,
) -> None:
"""Create a new experiment and save it if DBSettings available.
Args:
parameters: List of dictionaries representing parameters in the
experiment search space.
Required elements in the dictionaries are:
1. "name" (name of parameter, string),
2. "type" (type of parameter: "range", "fixed", or "choice", string),
and one of the following:
3a. "bounds" for range parameters (list of two values, lower bound
first),
3b. "values" for choice parameters (list of values), or
3c. "value" for fixed parameters (single value).
Optional elements are:
1. "log_scale" (for float-valued range parameters, bool),
2. "value_type" (to specify type that values of this parameter should
take; expects "float", "int", "bool" or "str"),
3. "is_fidelity" (bool) and "target_value" (float) for fidelity
parameters,
4. "is_ordered" (bool) for choice parameters, and
5. "is_task" (bool) for task parameters.
6. "digits" (int) for float-valued range parameters.
name: Name of the experiment to be created.
description: Description of the experiment to be created.
objective_name[DEPRECATED]: Name of the metric used as objective
in this experiment. This metric must be present in `raw_data`
argument to `complete_trial`.
minimize[DEPRECATED]: Whether this experiment represents a minimization
problem.
objectives: Mapping from an objective name to object containing:
minimize: Whether this experiment represents a minimization problem.
threshold: The bound in the objective's threshold constraint.
parameter_constraints: List of string representation of parameter
constraints, such as "x3 >= x4" or "-x3 + 2*x4 - 3.5*x5 >= 2". For
the latter constraints, any number of arguments is accepted, and
acceptable operators are "<=" and ">=". Note that parameter
constraints may only be placed on range parameters, not choice
parameters or fixed parameters.
outcome_constraints: List of string representation of outcome
constraints of form "metric_name >= bound", like "m1 <= 3."
status_quo: Parameterization of the current state of the system.
If set, this will be added to each trial to be evaluated alongside
test configurations.
overwrite_existing_experiment: If an experiment has already been set
on this `AxClient` instance, whether to reset it to the new one.
If overwriting the experiment, generation strategy will be
re-selected for the new experiment and restarted.
To protect experiments in production, one cannot overwrite existing
experiments if the experiment is already stored in the database,
regardless of the value of `overwrite_existing_experiment`.
tracking_metric_names: Names of additional tracking metrics not used for
optimization.
choose_generation_strategy_kwargs: Keyword arguments to pass to
`choose_generation_strategy` function which determines what
generation strategy should be used when none was specified on init.
support_intermediate_data: Whether trials may report intermediate results
for trials that are still running (i.e. have not been completed via
`ax_client.complete_trial`).
immutable_search_space_and_opt_config: Whether it's possible to update the
search space and optimization config on this experiment after creation.
Defaults to True. If set to True, we won't store or load copies of the
search space and optimization config on each generator run, which will
improve storage performance.
is_test: Whether this experiment will be a test experiment (useful for
marking test experiments in storage etc). Defaults to False.
metric_definitions: A mapping of metric names to extra kwargs to pass
to that metric
"""
self._validate_early_stopping_strategy(support_intermediate_data)
objective_kwargs = {}
if (objective_name or minimize is not None) and objectives:
raise UnsupportedError(
"You may either pass an an objective object "
"or an objective_name and minimize param, but not both"
)
elif objectives is not None:
objective_kwargs["objectives"] = {
objective: ("minimize" if properties.minimize else "maximize")
for objective, properties in objectives.items()
}
if len(objectives.keys()) > 1:
objective_kwargs[
"objective_thresholds"
] = self.build_objective_thresholds(objectives)
elif objective_name or minimize is not None:
objective_kwargs["objective_name"] = objective_name
objective_kwargs["minimize"] = minimize or False
warnings.warn(
"objective_name and minimize are deprecated",
category=DeprecationWarning,
)
experiment = self.make_experiment(
name=name,
description=description,
owners=owners,
parameters=parameters,
parameter_constraints=parameter_constraints,
outcome_constraints=outcome_constraints,
status_quo=status_quo,
experiment_type=experiment_type,
tracking_metric_names=tracking_metric_names,
metric_definitions=metric_definitions,
support_intermediate_data=support_intermediate_data,
immutable_search_space_and_opt_config=immutable_search_space_and_opt_config,
is_test=is_test,
**objective_kwargs,
)
self._set_runner(experiment=experiment)
self._set_experiment(
experiment=experiment,
overwrite_existing_experiment=overwrite_existing_experiment,
)
self._set_generation_strategy(
choose_generation_strategy_kwargs=choose_generation_strategy_kwargs
)
self._save_generation_strategy_to_db_if_possible()
@property
def status_quo(self) -> TParameterization:
"""The parameterization of the status quo arm of the experiment."""
return not_none(
self.experiment.status_quo, "Experiment does not have a status quo arm"
).parameters
[docs] def set_status_quo(self, params: Optional[TParameterization]) -> None:
"""Set, or unset status quo on the experiment. There may be risk
in using this after a trial with the status quo arm has run.
Args:
status_quo: Parameterization of the current state of the system.
If set, this will be added to each trial to be evaluated alongside
test configurations.
"""
self.experiment.status_quo = None if params is None else Arm(parameters=params)
[docs] def set_optimization_config(
self,
objectives: Optional[Dict[str, ObjectiveProperties]] = None,
outcome_constraints: Optional[List[str]] = None,
metric_definitions: Optional[Dict[str, Dict[str, Any]]] = None,
) -> None:
"""Overwrite experiment's optimization config
Args:
objectives: Mapping from an objective name to object containing:
minimize: Whether this experiment represents a minimization problem.
threshold: The bound in the objective's threshold constraint.
outcome_constraints: List of string representation of outcome
constraints of form "metric_name >= bound", like "m1 <= 3."
metric_definitions: A mapping of metric names to extra kwargs to pass
to that metric
"""
optimization_config = self.make_optimization_config_from_properties(
objectives=objectives,
outcome_constraints=outcome_constraints,
status_quo_defined=self.experiment.status_quo is not None,
metric_definitions=metric_definitions,
)
if optimization_config:
self.experiment.optimization_config = optimization_config
self._save_experiment_to_db_if_possible(
experiment=self.experiment,
)
else:
raise ValueError(
"optimization config not set because it was missing objectives"
)
[docs] def add_tracking_metrics(
self,
metric_names: List[str],
metric_definitions: Optional[Dict[str, Dict[str, Any]]] = None,
) -> None:
"""Add a list of new metrics to the experiment.
If any of the metrics are already defined on the experiment,
we raise an error and don't add any of them to the experiment
Args:
metric_names: Names of metrics to be added.
metric_definitions: A mapping of metric names to extra kwargs to pass
to that metric
"""
self.experiment.add_tracking_metrics(
metrics=[
self._make_metric(
name=metric_name, metric_definitions=metric_definitions
)
for metric_name in metric_names
]
)
[docs] @copy_doc(Experiment.remove_tracking_metric)
def remove_tracking_metric(self, metric_name: str) -> None:
self.experiment.remove_tracking_metric(metric_name=metric_name)
[docs] def set_search_space(
self,
parameters: List[
Dict[str, Union[TParamValue, Sequence[TParamValue], Dict[str, List[str]]]]
],
parameter_constraints: Optional[List[str]] = None,
) -> None:
"""Sets the search space on the experiment and saves.
This is expected to fail on base AxClient as experiment will have
immutable search space and optimization config set to True by default
Args:
parameters: List of dictionaries representing parameters in the
experiment search space.
Required elements in the dictionaries are:
1. "name" (name of parameter, string),
2. "type" (type of parameter: "range", "fixed", or "choice", string),
and one of the following:
3a. "bounds" for range parameters (list of two values, lower bound
first),
3b. "values" for choice parameters (list of values), or
3c. "value" for fixed parameters (single value).
Optional elements are:
1. "log_scale" (for float-valued range parameters, bool),
2. "value_type" (to specify type that values of this parameter should
take; expects "float", "int", "bool" or "str"),
3. "is_fidelity" (bool) and "target_value" (float) for fidelity
parameters,
4. "is_ordered" (bool) for choice parameters, and
5. "is_task" (bool) for task parameters.
6. "digits" (int) for float-valued range parameters.
parameter_constraints: List of string representation of parameter
constraints, such as "x3 >= x4" or "-x3 + 2*x4 - 3.5*x5 >= 2". For
the latter constraints, any number of arguments is accepted, and
acceptable operators are "<=" and ">=". Note that parameter
constraints may only be placed on range parameters, not choice
parameters or fixed parameters.
"""
self.experiment.search_space = self.make_search_space(
parameters=parameters, parameter_constraints=parameter_constraints
)
self._save_experiment_to_db_if_possible(
experiment=self.experiment,
)
[docs] @retry_on_exception(
logger=logger,
exception_types=(RuntimeError,),
check_message_contains=["Cholesky", "cholesky"],
suppress_all_errors=False,
wrap_error_message_in=CHOLESKY_ERROR_ANNOTATION,
)
def get_next_trial(
self, ttl_seconds: Optional[int] = None, force: bool = False
) -> Tuple[TParameterization, int]:
"""
Generate trial with the next set of parameters to try in the iteration process.
Note: Service API currently supports only 1-arm trials.
Args:
ttl_seconds: If specified, will consider the trial failed after this
many seconds. Used to detect dead trials that were not marked
failed properly.
force: If set to True, this function will bypass the global stopping
strategy's decision and generate a new trial anyway.
Returns:
Tuple of trial parameterization, trial index
"""
# Check if the global stopping strategy suggests to stop the optimization.
# This is needed only if there is actually a stopping strategy specified,
# and if this function is not forced to generate a new trial.
if self.global_stopping_strategy and (not force):
# The strategy itself will check if enough trials have already been
# completed.
(
stop_optimization,
global_stopping_message,
) = self.global_stopping_strategy.should_stop_optimization(
experiment=self.experiment
)
if stop_optimization:
raise OptimizationShouldStop(message=global_stopping_message)
try:
trial = self.experiment.new_trial(
generator_run=self._gen_new_generator_run(), ttl_seconds=ttl_seconds
)
except MaxParallelismReachedException as e:
if self._early_stopping_strategy is not None:
e.message += ( # noqa: B306
" When stopping trials early, make sure to call `stop_trial_early` "
"on the stopped trial."
)
raise e
logger.info(
f"Generated new trial {trial.index} with parameters "
f"{round_floats_for_logging(item=not_none(trial.arm).parameters)}."
)
trial.mark_running(no_runner_required=True)
self._save_or_update_trial_in_db_if_possible(
experiment=self.experiment,
trial=trial,
)
# TODO[T79183560]: Ensure correct handling of generator run when using
# foreign keys.
self._update_generation_strategy_in_db_if_possible(
generation_strategy=self.generation_strategy,
new_generator_runs=[self.generation_strategy._generator_runs[-1]],
)
return not_none(trial.arm).parameters, trial.index
[docs] def get_current_trial_generation_limit(self) -> Tuple[int, bool]:
"""How many trials this ``AxClient`` instance can currently produce via
calls to ``get_next_trial``, before more trials are completed, and whether
the optimization is complete.
NOTE: If return value of this function is ``(0, False)``, no more trials
can currently be procuded by this ``AxClient`` instance, but optimization
is not completed; once more trials are completed with data, more new
trials can be generated.
Returns: a two-item tuple of:
- the number of trials that can currently be produced, with -1
meaning unlimited trials,
- whether no more trials can be produced by this ``AxClient``
instance at any point (e.g. if the search space is exhausted or
generation strategy is completed.
"""
# Ensure that experiment is set on the generation strategy.
if self.generation_strategy._experiment is None:
self.generation_strategy.experiment = self.experiment
return self.generation_strategy.current_generator_run_limit()
[docs] def get_next_trials(
self, max_trials: int, ttl_seconds: Optional[int] = None
) -> Tuple[Dict[int, TParameterization], bool]:
"""Generate as many trials as currently possible.
NOTE: Useful for running multiple trials in parallel: produces multiple trials,
with their number limited by:
- parallelism limit on current generation step,
- number of trials in current generation step,
- number of trials required to complete before moving to next generation step,
if applicable,
- and ``max_trials`` argument to this method.
Args:
max_trials: Limit on how many trials the call to this method should produce.
ttl_seconds: If specified, will consider the trial failed after this
many seconds. Used to detect dead trials that were not marked
failed properly.
Returns: two-item tuple of:
- mapping from trial indices to parameterizations in those trials,
- boolean indicator of whether optimization is completed and no more
trials can be generated going forward.
"""
gen_limit, optimization_complete = self.get_current_trial_generation_limit()
if optimization_complete:
return {}, True
# Trial generation limit of -1 indicates that unlimited trials can be
# generated, so we only want to limit `max_trials` if `trial_generation_
# limit` is non-negative.
if gen_limit >= 0:
max_trials = min(gen_limit, max_trials)
trials_dict = {}
for _ in range(max_trials):
try:
params, trial_index = self.get_next_trial(ttl_seconds=ttl_seconds)
trials_dict[trial_index] = params
except OptimizationComplete as err:
logger.info(
f"Encountered exception indicating optimization completion: {err}"
)
return trials_dict, True
# Check whether optimization is complete now that we generated a batch
# of trials.
_, optimization_complete = self.get_current_trial_generation_limit()
return trials_dict, optimization_complete
[docs] def abandon_trial(self, trial_index: int, reason: Optional[str] = None) -> None:
"""Abandons a trial and adds optional metadata to it.
Args:
trial_index: Index of trial within the experiment.
"""
trial = self.get_trial(trial_index)
trial.mark_abandoned(reason=reason)
[docs] def complete_trial(
self,
trial_index: int,
raw_data: TEvaluationOutcome,
metadata: Optional[Dict[str, Union[str, int]]] = None,
sample_size: Optional[int] = None,
) -> None:
"""
Completes the trial with given metric values and adds optional metadata
to it.
NOTE: When ``raw_data`` does not specify SEM for a given metric, Ax
will default to the assumption that the data is noisy (specifically,
corrupted by additive zero-mean Gaussian noise) and that the
level of noise should be inferred by the optimization model. To
indicate that the data is noiseless, set SEM to 0.0, for example:
.. code-block:: python
ax_client.complete_trial(
trial_index=0,
raw_data={"my_objective": (objective_mean_value, 0.0)}
)
Args:
trial_index: Index of trial within the experiment.
raw_data: Evaluation data for the trial. Can be a mapping from
metric name to a tuple of mean and SEM, just a tuple of mean and
SEM if only one metric in optimization, or just the mean if SEM is
unknown (then Ax will infer observation noise level).
Can also be a list of (fidelities, mapping from
metric name to a tuple of mean and SEM).
metadata: Additional metadata to track about this run.
sample_size: Number of samples collected for the underlying arm,
optional.
"""
# Validate that trial can be completed.
trial = self.get_trial(trial_index)
self._validate_can_complete_trial(trial=trial)
if not isinstance(trial_index, int): # pragma: no cover
raise ValueError(f"Trial index must be an int, got: {trial_index}.")
data_update_repr = self._update_trial_with_raw_data(
trial_index=trial_index,
raw_data=raw_data,
metadata=metadata,
sample_size=sample_size,
complete_trial=True,
combine_with_last_data=True,
)
logger.info(f"Completed trial {trial_index} with data: " f"{data_update_repr}.")
[docs] def update_trial_data(
self,
trial_index: int,
raw_data: TEvaluationOutcome,
metadata: Optional[Dict[str, Union[str, int]]] = None,
sample_size: Optional[int] = None,
) -> None:
"""
Attaches additional data for completed trial (for example, if trial was
completed with data for only one of the required metrics and more data
needs to be attached).
Args:
trial_index: Index of trial within the experiment.
raw_data: Evaluation data for the trial. Can be a mapping from
metric name to a tuple of mean and SEM, just a tuple of mean and
SEM if only one metric in optimization, or just the mean if there
is no SEM. Can also be a list of (fidelities, mapping from
metric name to a tuple of mean and SEM).
metadata: Additional metadata to track about this run.
sample_size: Number of samples collected for the underlying arm,
optional.
"""
if not isinstance(trial_index, int): # pragma: no cover
raise ValueError(f"Trial index must be an int, got: {trial_index}.")
trial = self.get_trial(trial_index)
if not trial.status.is_completed:
raise ValueError(
f"Trial {trial.index} has not yet been completed with data."
"To complete it, use `ax_client.complete_trial`."
)
data_update_repr = self._update_trial_with_raw_data(
trial_index=trial_index,
raw_data=raw_data,
metadata=metadata,
sample_size=sample_size,
combine_with_last_data=True,
)
# Registering trial data update is needed for generation strategies that
# leverage the `update` functionality of model and bridge setup and therefore
# need to be aware of new data added to experiment. Usually this happends
# seamlessly, by looking at newly completed trials, but in this case trial
# status does not change, so we manually register the new data.
# Currently this call will only result in a `NotImplementedError` if generation
# strategy uses `update` (`GenerationStep.use_update` is False by default).
self.generation_strategy._register_trial_data_update(trial=trial)
logger.info(f"Added data: {data_update_repr} to trial {trial.index}.")
[docs] def log_trial_failure(
self, trial_index: int, metadata: Optional[Dict[str, str]] = None
) -> None:
"""Mark that the given trial has failed while running.
Args:
trial_index: Index of trial within the experiment.
metadata: Additional metadata to track about this run.
"""
trial = self.experiment.trials[trial_index]
trial.mark_failed()
logger.info(f"Registered failure of trial {trial_index}.")
if metadata is not None:
trial._run_metadata = metadata
self._save_experiment_to_db_if_possible(
experiment=self.experiment,
)
[docs] def attach_trial(
self,
parameters: TParameterization,
ttl_seconds: Optional[int] = None,
run_metadata: Optional[Dict[str, Any]] = None,
arm_name: Optional[str] = None,
) -> Tuple[TParameterization, int]:
"""Attach a new trial with the given parameterization to the experiment.
Args:
parameters: Parameterization of the new trial.
ttl_seconds: If specified, will consider the trial failed after this
many seconds. Used to detect dead trials that were not marked
failed properly.
Returns:
Tuple of parameterization and trial index from newly created trial.
"""
self._validate_search_space_membership(parameters=parameters)
# If search space is hierarchical, we need to store dummy values of parameters
# that are not in the arm (but are in flattened search space), as metadata,
# so later we are able to make the data for this arm "complete" in the
# flattened search space.
candidate_metadata = None
if self.experiment.search_space.is_hierarchical:
hss = checked_cast(HierarchicalSearchSpace, self.experiment.search_space)
candidate_metadata = hss.cast_observation_features(
observation_features=hss.flatten_observation_features(
observation_features=ObservationFeatures(parameters=parameters),
inject_dummy_values_to_complete_flat_parameterization=True,
)
).metadata
trial = self.experiment.new_trial(ttl_seconds=ttl_seconds).add_arm(
Arm(parameters=parameters, name=arm_name),
candidate_metadata=candidate_metadata,
)
trial.mark_running(no_runner_required=True)
logger.info(
"Attached custom parameterization "
f"{round_floats_for_logging(item=parameters)} as trial {trial.index}."
)
if run_metadata is not None:
trial.update_run_metadata(metadata=run_metadata)
self._save_or_update_trial_in_db_if_possible(
experiment=self.experiment,
trial=trial,
)
return not_none(trial.arm).parameters, trial.index
[docs] def get_trial_parameters(self, trial_index: int) -> TParameterization:
"""Retrieve the parameterization of the trial by the given index."""
return not_none(self.get_trial(trial_index).arm).parameters
[docs] def get_trials_data_frame(self) -> pd.DataFrame:
return exp_to_df(exp=self.experiment)
[docs] def get_max_parallelism(self) -> List[Tuple[int, int]]:
"""Retrieves maximum number of trials that can be scheduled in parallel
at different stages of optimization.
Some optimization algorithms profit significantly from sequential
optimization (i.e. suggest a few points, get updated with data for them,
repeat, see https://ax.dev/docs/bayesopt.html).
Parallelism setting indicates how many trials should be running simulteneously
(generated, but not yet completed with data).
The output of this method is mapping of form
{num_trials -> max_parallelism_setting}, where the max_parallelism_setting
is used for num_trials trials. If max_parallelism_setting is -1, as
many of the trials can be ran in parallel, as necessary. If num_trials
in a tuple is -1, then the corresponding max_parallelism_setting
should be used for all subsequent trials.
For example, if the returned list is [(5, -1), (12, 6), (-1, 3)],
the schedule could be: run 5 trials with any parallelism, run 6 trials in
parallel twice, run 3 trials in parallel for as long as needed. Here,
'running' a trial means obtaining a next trial from `AxClient` through
get_next_trials and completing it with data when available.
Returns:
Mapping of form {num_trials -> max_parallelism_setting}.
"""
parallelism_settings = []
for step in self.generation_strategy._steps:
parallelism_settings.append(
(step.num_trials, step.max_parallelism or step.num_trials)
)
return parallelism_settings
[docs] def get_optimization_trace(
self, objective_optimum: Optional[float] = None
) -> AxPlotConfig:
"""Retrieves the plot configuration for optimization trace, which shows
the evolution of the objective mean over iterations.
Args:
objective_optimum: Optimal objective, if known, for display in the
visualization.
"""
if not self.experiment.trials:
raise ValueError("Cannot generate plot as there are no trials.")
objective = self.objective
if isinstance(objective, MultiObjective):
raise UnsupportedError(
"`get_optimization_trace` is not supported "
"for multi-objective experiments"
)
objective_name = self.objective_name
best_objectives = np.array(
[
[
checked_cast(Trial, trial).objective_mean
for trial in self.experiment.trials.values()
if trial.status.is_completed
]
]
)
hover_labels = [
_format_dict(not_none(checked_cast(Trial, trial).arm).parameters)
for trial in self.experiment.trials.values()
if trial.status.is_completed
]
return optimization_trace_single_method(
y=(
np.minimum.accumulate(best_objectives, axis=1)
if objective.minimize
else np.maximum.accumulate(best_objectives, axis=1)
),
optimum=objective_optimum,
title="Model performance vs. # of iterations",
ylabel=objective_name.capitalize(),
hover_labels=hover_labels,
model_transitions=self.generation_strategy.model_transitions,
)
[docs] def get_contour_plot(
self,
param_x: Optional[str] = None,
param_y: Optional[str] = None,
metric_name: Optional[str] = None,
) -> AxPlotConfig:
"""Retrieves a plot configuration for a contour plot of the response
surface. For response surfaces with more than two parameters,
selected two parameters will appear on the axes, and remaining parameters
will be affixed to the middle of their range. If contour params arguments
are not provided, the first two parameters in the search space will be
used. If contour metrics are not provided, objective will be used.
Args:
param_x: name of parameters to use on x-axis for
the contour response surface plots.
param_y: name of parameters to use on y-axis for
the contour response surface plots.
metric_name: Name of the metric, for which to plot the response
surface.
"""
if not self.experiment.trials:
raise ValueError("Cannot generate plot as there are no trials.")
if len(self.experiment.parameters) < 2:
raise ValueError(
"Cannot create a contour plot as experiment has less than 2 "
"parameters, but a contour-related argument was provided."
)
if (param_x or param_y) and not (param_x and param_y):
raise ValueError(
"If `param_x` is provided, `param_y` is "
"required as well, and vice-versa."
)
if not metric_name:
if isinstance(self.objective, MultiObjective):
raise UnsupportedError(
"`get_contour_plot` requires a `metric_name` "
"for multi-objective experiments"
)
metric_name = self.objective_name
if not param_x or not param_y:
parameter_names = list(self.experiment.parameters.keys())
param_x = parameter_names[0]
param_y = parameter_names[1]
if param_x not in self.experiment.parameters:
raise ValueError(
f'Parameter "{param_x}" not found in the optimization search space.'
)
if param_y not in self.experiment.parameters:
raise ValueError(
f'Parameter "{param_y}" not found in the optimization search space.'
)
if metric_name not in self.experiment.metrics:
raise ValueError(
f'Metric "{metric_name}" is not associated with this optimization.'
)
if self.generation_strategy.model is not None:
try:
logger.info(
f"Retrieving contour plot with parameter '{param_x}' on X-axis "
f"and '{param_y}' on Y-axis, for metric '{metric_name}'. "
"Remaining parameters are affixed to the middle of their range."
)
return plot_contour(
model=not_none(self.generation_strategy.model),
param_x=param_x,
param_y=param_y,
metric_name=metric_name,
)
except NotImplementedError:
# Some models don't implement '_predict', which is needed
# for the contour plots.
logger.info(
f"Model {self.generation_strategy.model} does not implement "
"`predict`, so it cannot be used to generate a response "
"surface plot."
)
raise UnsupportedPlotError(
f'Could not obtain contour plot of "{metric_name}" for parameters '
f'"{param_x}" and "{param_y}", as a model with predictive ability, '
"such as a Gaussian Process, has not yet been trained in the course "
"of this optimization."
)
[docs] def get_feature_importances(self, relative: bool = True) -> AxPlotConfig:
"""
Get a bar chart showing feature_importances for a metric.
A drop-down controls the metric for which the importances are displayed.
Args:
relative: Whether the values are displayed as percentiles or
as raw importance metrics.
"""
if not self.experiment.trials:
raise ValueError("Cannot generate plot as there are no trials.")
cur_model = self.generation_strategy.model
if cur_model is not None:
try:
return plot_feature_importance_by_feature(cur_model, relative=relative)
except NotImplementedError:
logger.info(
f"Model {self.generation_strategy.model} does not implement "
"`feature_importances`, so it cannot be used to generate "
"this plot. Only certain models, specifically GPEI, implement "
"feature importances."
)
raise ValueError(
"Could not obtain feature_importances for any metrics "
" as a model that can produce feature importances, such as a "
"Gaussian Process, has not yet been trained in the course "
"of this optimization."
)
[docs] def load_experiment_from_database(
self,
experiment_name: str,
choose_generation_strategy_kwargs: Optional[Dict[str, Any]] = None,
) -> None:
"""Load an existing experiment from database using the `DBSettings`
passed to this `AxClient` on instantiation.
Args:
experiment_name: Name of the experiment.
Returns:
Experiment object.
"""
experiment, generation_strategy = self._load_experiment_and_generation_strategy(
experiment_name=experiment_name
)
self._experiment = not_none(
experiment, f"Experiment by name '{experiment_name}' not found."
)
logger.info(f"Loaded {experiment}.")
if generation_strategy is None: # pragma: no cover
self._set_generation_strategy(
choose_generation_strategy_kwargs=choose_generation_strategy_kwargs
)
self._save_generation_strategy_to_db_if_possible()
else:
self._generation_strategy = generation_strategy
logger.info(
f"Using generation strategy associated with the loaded experiment:"
f" {generation_strategy}."
)
[docs] def get_model_predictions_for_parameterizations(
self,
parameterizations: List[TParameterization],
metric_names: Optional[List[str]] = None,
) -> List[Dict[str, Tuple[float, float]]]:
"""Retrieve model-estimated means and covariances for all metrics
for the provided parameterizations.
Args:
metric_names: Names of the metrics for which to predict.
All metrics will be predicted if this argument is
not specified.
parameterizations: List of Parameterizations for which to predict.
Returns:
A list of predicted metric mean and SEM of form:
List[Tuple[float, float]].
"""
parameterization_dict = {
i: parameterization for i, parameterization in enumerate(parameterizations)
}
predictions_dict = self.get_model_predictions(
metric_names=metric_names, parameterizations=parameterization_dict
)
predictions_array = [
predictions_dict[i] for i, _ in enumerate(parameterizations)
]
return predictions_array
[docs] def get_model_predictions(
self,
metric_names: Optional[List[str]] = None,
include_out_of_sample: Optional[bool] = True,
parameterizations: Optional[Dict[int, TParameterization]] = None,
) -> Dict[int, Dict[str, Tuple[float, float]]]:
"""Retrieve model-estimated means and covariances for all metrics.
Args:
metric_names: Names of the metrics, for which to retrieve predictions.
All metrics on experiment will be retrieved if this argument was
not specified.
include_out_of_sample: Defaults to True. Return predictions for
out-of-sample (i.e. not yet completed trials) data in
addition to in-sample (i.e. completed trials) data.
parameterizations: Optional mapping from an int label to
Parameterizations. When provided, predictions are performed *only*
on these data points, no predictions from trial data is performed,
and include_out_of_sample parameters is ignored.
Returns:
A mapping from trial index to a mapping of metric names to tuples
of predicted metric mean and SEM, of form:
{ trial_index -> { metric_name: ( mean, SEM ) } }.
Note that AxClient currently support only 1-arm trials. i.e.
trial_index describes the single arms attached to the referenced
trial.
"""
# Ensure there are metrics specified
if metric_names is None and self.experiment.metrics is None:
raise ValueError( # pragma: no cover
"No metrics to retrieve specified on the experiment or as "
"argument to `get_model_predictions`."
)
# Fit model to ensure:
# - model is instantiated if needed
# - any new completed trials are fit to the model
self.fit_model()
# Shared info for subsequent calls
metric_names_to_predict = (
set(metric_names)
if metric_names is not None
else set(not_none(self.experiment.metrics).keys())
)
model = not_none(
self.generation_strategy.model, "No model has been instantiated yet."
)
# Construct a dictionary that maps from a label to an
# ObservationFeature to predict.
# - If returning trial predictions, the label is the trial index.
# - If predictions are for user-provided parameterization, the label
# is provided in the input (also an int).
label_to_feature_dict = {}
# Predict on user-provided data
if parameterizations is not None:
logger.info(
'"parameterizations" have been provided, only these data '
"points will be predicted. No trial data prediction will be "
"returned."
)
for label in parameterizations.keys():
label_to_feature_dict[label] = ObservationFeatures(
parameters=parameterizations[label]
)
# Predict on associated trials
else:
# Note that currently AxClient supports only 1-arm trials.
trials_dict = self.experiment.trials
for trial_index, trial in trials_dict.items():
# filter trials based on input params and trial statuses
if include_out_of_sample or trial.status.is_completed:
arms = trial.arms
if len(arms) > 1:
raise ValueError("Currently only 1-arm trials are supported.")
label_to_feature_dict[trial_index] = ObservationFeatures.from_arm(
arms[0]
)
return predict_by_features(
model=model,
label_to_feature_dict=label_to_feature_dict,
metric_names=metric_names_to_predict,
)
[docs] def fit_model(self) -> None:
"""Fit any completed trial data to the model.
If no model is yet available a new one is instantiated. This
may be the case when get_next_trial() has never been called.
"""
# Try to instantiate a model if there is none. This
# handles the case where trials have not be generated by calling
# get_next_trial().
if self.generation_strategy.model is None:
logger.info(
"get_model_predictions() has been called when no model is "
"instantiated. Attempting to instantiate the model for the "
"first time."
)
if not self.experiment.trial_indices_by_status[TrialStatus.COMPLETED]:
raise ValueError(
"At least one trial must be completed with data to instantiate "
"a model."
)
self.generation_strategy._fit_or_update_current_model(data=None)
logger.info("Successfully instantiated a model for the first time.")
# Model update is normally tied to the GenerationStrategy.gen() call,
# which is called from get_next_trial(). In order to ensure that predictions
# can be performed without the need to call get_next_trial(), we update the
# model with all attached data. Note that this method keeps track of previously
# seen trials and will update the model if there is newly attached data.
self.generation_strategy._fit_or_update_current_model(data=None)
[docs] def verify_trial_parameterization(
self, trial_index: int, parameterization: TParameterization
) -> bool:
"""Whether the given parameterization matches that of the arm in the trial
specified in the trial index.
"""
return not_none(self.get_trial(trial_index).arm).parameters == parameterization
[docs] def should_stop_trials_early(
self, trial_indices: Set[int]
) -> Dict[int, Optional[str]]:
"""Evaluate whether to early-stop running trials.
Args:
trial_indices: Indices of trials to consider for early stopping.
Returns:
A dictionary mapping trial indices that should be early stopped to
(optional) messages with the associated reason.
"""
if self._early_stopping_strategy is None:
logger.warning(
"No early_stopping_strategy was passed to AxClient. "
"Defaulting to never stopping any trials early."
)
return early_stopping_utils.should_stop_trials_early(
early_stopping_strategy=self._early_stopping_strategy,
trial_indices=trial_indices,
experiment=self.experiment,
)
[docs] def stop_trial_early(self, trial_index: int) -> None:
trial = self.get_trial(trial_index)
trial.mark_early_stopped()
logger.info(f"Early stopped trial {trial_index}.")
# ------------------ JSON serialization & storage methods. -----------------
[docs] def save_to_json_file(self, filepath: str = "ax_client_snapshot.json") -> None:
"""Save a JSON-serialized snapshot of this `AxClient`'s settings and state
to a .json file by the given path.
"""
with open(filepath, "w+") as file: # pragma: no cover
file.write(json.dumps(self.to_json_snapshot()))
logger.info(f"Saved JSON-serialized state of optimization to `{filepath}`.")
[docs] @classmethod
def load_from_json_file(
cls: Type[AxClientSubclass],
filepath: str = "ax_client_snapshot.json",
# pyre-fixme[2]: Parameter must be annotated.
**kwargs,
) -> AxClientSubclass:
"""Restore an `AxClient` and its state from a JSON-serialized snapshot,
residing in a .json file by the given path.
"""
with open(filepath, "r") as file: # pragma: no cover
serialized = json.loads(file.read())
return cls.from_json_snapshot(serialized=serialized, **kwargs)
[docs] def to_json_snapshot(
self,
# pyre-fixme[2]: Parameter annotation cannot contain `Any`.
# pyre-fixme[24]: Generic type `type` expects 1 type parameter, use
# `typing.Type` to avoid runtime subscripting errors.
encoder_registry: Optional[Dict[Type, Callable[[Any], Dict[str, Any]]]] = None,
# pyre-fixme[2]: Parameter annotation cannot contain `Any`.
# pyre-fixme[24]: Generic type `type` expects 1 type parameter, use
# `typing.Type` to avoid runtime subscripting errors.
class_encoder_registry: Optional[
Dict[Type, Callable[[Any], Dict[str, Any]]]
] = None,
) -> Dict[str, Any]:
"""Serialize this `AxClient` to JSON to be able to interrupt and restart
optimization and save it to file by the provided path.
Returns:
A JSON-safe dict representation of this `AxClient`.
"""
if encoder_registry is None:
encoder_registry = CORE_ENCODER_REGISTRY
if class_encoder_registry is None:
class_encoder_registry = CORE_CLASS_ENCODER_REGISTRY
return {
"_type": self.__class__.__name__,
"experiment": object_to_json(
self._experiment,
encoder_registry=encoder_registry,
class_encoder_registry=class_encoder_registry,
),
"generation_strategy": object_to_json(
self._generation_strategy,
encoder_registry=encoder_registry,
class_encoder_registry=class_encoder_registry,
),
"_enforce_sequential_optimization": self._enforce_sequential_optimization,
}
[docs] @classmethod
def from_json_snapshot(
cls: Type[AxClientSubclass],
serialized: Dict[str, Any],
# pyre-fixme[24]: Generic type `type` expects 1 type parameter, use
# `typing.Type` to avoid runtime subscripting errors.
decoder_registry: Optional[Dict[str, Type]] = None,
# pyre-fixme[2]: Parameter annotation cannot contain `Any`.
class_decoder_registry: Optional[
Dict[str, Callable[[Dict[str, Any]], Any]]
] = None,
# pyre-fixme[2]: Parameter must be annotated.
**kwargs,
) -> AxClientSubclass:
"""Recreate an `AxClient` from a JSON snapshot."""
if decoder_registry is None:
decoder_registry = CORE_DECODER_REGISTRY
if class_decoder_registry is None:
class_decoder_registry = CORE_CLASS_DECODER_REGISTRY
experiment = object_from_json(
serialized.pop("experiment"),
decoder_registry=decoder_registry,
class_decoder_registry=class_decoder_registry,
)
serialized_generation_strategy = serialized.pop("generation_strategy")
ax_client = cls(
generation_strategy=generation_strategy_from_json(
generation_strategy_json=serialized_generation_strategy,
experiment=experiment,
decoder_registry=decoder_registry,
class_decoder_registry=class_decoder_registry,
)
if serialized_generation_strategy is not None
else None,
enforce_sequential_optimization=serialized.pop(
"_enforce_sequential_optimization"
),
**kwargs,
)
ax_client._experiment = experiment
return ax_client
# ---------------------- Private helper methods. ---------------------
@property
def experiment(self) -> Experiment:
"""Returns the experiment set on this Ax client."""
return not_none(
self._experiment,
(
"Experiment not set on Ax client. Must first "
"call load_experiment or create_experiment to use handler functions."
),
)
[docs] def get_trial(self, trial_index: int) -> Trial:
"""Return a trial on experiment cast as Trial"""
return checked_cast(Trial, self.experiment.trials[trial_index])
@property
def generation_strategy(self) -> GenerationStrategy:
"""Returns the generation strategy, set on this experiment."""
return not_none(
self._generation_strategy,
"No generation strategy has been set on this optimization yet.",
)
@property
def objective(self) -> Objective:
return not_none(self.experiment.optimization_config).objective
@property
def objective_name(self) -> str:
"""Returns the name of the objective in this optimization."""
objective = self.objective
if isinstance(objective, MultiObjective):
raise UnsupportedError(
"Multi-objective experiments contain multiple objectives"
)
return objective.metric.name
@property
def objective_names(self) -> List[str]:
"""Returns the name of the objective in this optimization."""
objective = self.objective
return [m.name for m in objective.metrics]
@property
def metric_definitions(self) -> Dict[str, Dict[str, Any]]:
"""Returns metric definitions for all experiment metrics that can
be passed into functions requiring metric_definitions
"""
return {
m.serialize_init_args(m)["name"]: {
"metric_class": m.__class__,
**{k: v for k, v in m.serialize_init_args(m).items() if k != "name"},
}
for m in self.experiment.metrics.values()
}
@property
def metric_names(self) -> Set[str]:
"""Returns the names of all metrics on the attached experiment."""
return set(self.experiment.metrics)
@property
def global_stopping_strategy(self) -> Optional[BaseGlobalStoppingStrategy]:
"""The global stopping strategy used on the experiment."""
return self._global_stopping_strategy
@global_stopping_strategy.setter
def global_stopping_strategy(self, gss: BaseGlobalStoppingStrategy) -> None:
"""Update the global stopping strategy."""
self._global_stopping_strategy = gss
[docs] @copy_doc(BestPointMixin.get_best_trial)
def get_best_trial(
self,
optimization_config: Optional[OptimizationConfig] = None,
trial_indices: Optional[Iterable[int]] = None,
use_model_predictions: bool = True,
) -> Optional[Tuple[int, TParameterization, Optional[TModelPredictArm]]]:
return self._get_best_trial(
experiment=self.experiment,
generation_strategy=self.generation_strategy,
trial_indices=trial_indices,
use_model_predictions=use_model_predictions,
)
[docs] @copy_doc(BestPointMixin.get_pareto_optimal_parameters)
def get_pareto_optimal_parameters(
self,
optimization_config: Optional[OptimizationConfig] = None,
trial_indices: Optional[Iterable[int]] = None,
use_model_predictions: bool = True,
) -> Dict[int, Tuple[TParameterization, TModelPredictArm]]:
return self._get_pareto_optimal_parameters(
experiment=self.experiment,
generation_strategy=self.generation_strategy,
trial_indices=trial_indices,
use_model_predictions=use_model_predictions,
)
[docs] @copy_doc(BestPointMixin.get_hypervolume)
def get_hypervolume(
self,
optimization_config: Optional[MultiObjectiveOptimizationConfig] = None,
trial_indices: Optional[Iterable[int]] = None,
use_model_predictions: bool = True,
) -> float:
return BestPointMixin._get_hypervolume(
experiment=self.experiment,
generation_strategy=self.generation_strategy,
optimization_config=optimization_config,
trial_indices=trial_indices,
use_model_predictions=use_model_predictions,
)
def _update_trial_with_raw_data(
self,
trial_index: int,
raw_data: TEvaluationOutcome,
metadata: Optional[Dict[str, Union[str, int]]] = None,
sample_size: Optional[int] = None,
complete_trial: bool = False,
combine_with_last_data: bool = False,
) -> str:
"""Helper method attaches data to a trial, returns a str of update."""
# Format the data to save.
trial = self.get_trial(trial_index)
sample_sizes = {not_none(trial.arm).name: sample_size} if sample_size else {}
evaluations, data = self._make_evaluations_and_data(
trial=trial, raw_data=raw_data, metadata=metadata, sample_sizes=sample_sizes
)
metadata = metadata or {}
self._validate_trial_data(trial=trial, data=data)
trial.update_run_metadata(metadata=metadata)
self.experiment.attach_data(
data=data, combine_with_last_data=combine_with_last_data
)
if complete_trial:
trial.mark_completed()
self._save_or_update_trial_in_db_if_possible(
experiment=self.experiment,
trial=trial,
)
return str(
round_floats_for_logging(item=evaluations[next(iter(evaluations.keys()))])
)
def _set_experiment(
self,
experiment: Experiment,
overwrite_existing_experiment: bool = False,
) -> None:
"""Sets the ``_experiment`` attribute on this `AxClient`` instance and saves the
experiment if this instance uses SQL storage.
NOTE: This setter **should not be used outside of this file in production**.
It can be leveraged in development, but all checked-in code that uses the
Service API should leverage ``AxClient.create_experiment`` instead and extend it
as needed. If using ``create_experiment`` is impossible and this setter is
required, please raise your use case to the AE team or on our Github.
"""
name = experiment._name
if self.db_settings_set and not name:
raise ValueError( # pragma: no cover
"Must give the experiment a name if `db_settings` is not None."
)
if self.db_settings_set:
experiment_id, _ = self._get_experiment_and_generation_strategy_db_id(
experiment_name=not_none(name)
)
if experiment_id:
raise ValueError(
f"Experiment {name} already exists in the database. "
"To protect experiments that are running in production, "
"overwriting stored experiments is not allowed. To "
"start a new experiment and store it, change the "
"experiment's name."
)
if self._experiment is not None:
if overwrite_existing_experiment:
exp_name = self.experiment._name or "untitled"
new_exp_name = name or "untitled"
logger.info(
f"Overwriting existing experiment ({exp_name}) on this client "
f"with new experiment ({new_exp_name}) and restarting the "
"generation strategy."
)
self._generation_strategy = None
else:
raise ValueError(
"Experiment already created for this client instance. "
"Set the `overwrite_existing_experiment` to `True` to overwrite "
"with new experiment."
)
self._experiment = experiment
try:
self._save_experiment_to_db_if_possible(
experiment=self.experiment,
)
except Exception:
# Unset the experiment on this `AxClient` instance if encountered and
# raising an error from saving the experiment, to avoid a case where
# overall `create_experiment` call fails with a storage error, but
# `self._experiment` is still set and user has to specify the
# `overwrite_existing_experiment` kwarg to re-attempt exp. creation.
self._experiment = None
raise
def _set_runner(self, experiment: Experiment) -> None:
"""Overridable method to sets a runner on the experiment."""
experiment.runner = None
def _set_generation_strategy(
self, choose_generation_strategy_kwargs: Optional[Dict[str, Any]] = None
) -> None:
"""Selects the generation strategy and applies specified dispatch kwargs,
if any.
"""
choose_generation_strategy_kwargs = choose_generation_strategy_kwargs or {}
random_seed = choose_generation_strategy_kwargs.pop(
"random_seed", self._random_seed
)
enforce_sequential_optimization = choose_generation_strategy_kwargs.pop(
"enforce_sequential_optimization", self._enforce_sequential_optimization
)
if self._generation_strategy is None:
self._generation_strategy = choose_generation_strategy(
search_space=self.experiment.search_space,
optimization_config=self.experiment.optimization_config,
enforce_sequential_optimization=enforce_sequential_optimization,
random_seed=random_seed,
torch_device=self._torch_device,
experiment=self.experiment,
**choose_generation_strategy_kwargs,
)
elif self._experiment:
self._generation_strategy.experiment = self.experiment
def _save_generation_strategy_to_db_if_possible(
self,
generation_strategy: Optional[GenerationStrategy] = None,
suppress_all_errors: bool = False,
) -> bool:
return super()._save_generation_strategy_to_db_if_possible(
generation_strategy=generation_strategy or self.generation_strategy,
suppress_all_errors=suppress_all_errors,
)
def _gen_new_generator_run(self, n: int = 1) -> GeneratorRun:
"""Generate new generator run for this experiment.
Args:
n: Number of arms to generate.
"""
# If random seed is not set for this optimization, context manager does
# nothing; otherwise, it sets the random seed for torch, but only for the
# scope of this call. This is important because torch seed is set globally,
# so if we just set the seed without the context manager, it can have
# serious negative impact on the performance of the models that employ
# stochasticity.
with manual_seed(seed=self._random_seed):
return not_none(self.generation_strategy).gen(
experiment=self.experiment,
n=n,
pending_observations=self._get_pending_observation_features(
experiment=self.experiment
),
)
def _find_last_trial_with_parameterization(
self, parameterization: TParameterization
) -> int:
"""Given a parameterization, find the last trial in the experiment that
contains an arm with that parameterization.
"""
for trial_idx in sorted(self.experiment.trials.keys(), reverse=True):
if not_none(self.get_trial(trial_idx).arm).parameters == parameterization:
return trial_idx
raise ValueError(
f"No trial on experiment matches parameterization {parameterization}."
)
@classmethod
def _get_pending_observation_features(
cls,
# pyre-fixme[2]: Parameter must be annotated.
experiment,
) -> Optional[Dict[str, List[ObservationFeatures]]]:
"""Extract pending points for the given experiment.
NOTE: With one-arm `Trial`-s, we use a more performant
``get_pending_observation_features_based_on_trial_status`` utility instead
of ``get_pending_observation_features``, since we can determine whether a point
is pending based on the status of the corresponding trial.
"""
return get_pending_observation_features_based_on_trial_status(
experiment=experiment
)
@classmethod
def _raw_data_by_arm(
cls,
trial: BaseTrial,
raw_data: Union[TEvaluationOutcome, Dict[str, TEvaluationOutcome]],
) -> Dict[str, TEvaluationOutcome]:
raw_data_by_arm: Dict[str, TEvaluationOutcome]
if isinstance(trial, BatchTrial): # pragma: no cover
raw_data_by_arm = checked_cast_complex(
Dict[str, TEvaluationOutcome],
raw_data,
message=cls.BATCH_TRIAL_RAW_DATA_FORMAT_ERROR_MESSAGE,
)
elif isinstance(trial, Trial):
arm_name = not_none(trial.arm).name
raw_data_by_arm = {
arm_name: checked_cast_complex(
TEvaluationOutcome,
raw_data,
message=cls.TRIAL_RAW_DATA_FORMAT_ERROR_MESSAGE,
)
}
else: # pragma: no cover
raise ValueError(f"Unexpected trial type: {type(trial)}.")
not_trial_arm_names = set(raw_data_by_arm.keys()) - set(
trial.arms_by_name.keys()
)
if not_trial_arm_names:
raise ValueError( # pragma: no cover
f"Arms {not_trial_arm_names} are not part of trial #{trial.index}."
)
return raw_data_by_arm
def _make_evaluations_and_data(
self,
trial: BaseTrial,
raw_data: Union[TEvaluationOutcome, Dict[str, TEvaluationOutcome]],
metadata: Optional[Dict[str, Union[str, int]]],
sample_sizes: Optional[Dict[str, int]] = None,
) -> Tuple[Dict[str, TEvaluationOutcome], Data]:
"""Formats given raw data as Ax evaluations and `Data`.
Args:
trial: Trial within the experiment.
raw_data: Metric outcomes for 1-arm trials, map from arm name to
metric outcomes for batched trials.
sample_size: Integer sample size for 1-arm trials, dict from arm
name to sample size for batched trials. Optional.
metadata: Additional metadata to track about this run.
data_is_for_batched_trials: Whether making evaluations and data for
a batched trial or a 1-arm trial.
"""
raw_data_by_arm = self._raw_data_by_arm(trial=trial, raw_data=raw_data)
metadata = metadata if metadata is not None else {}
evaluations, data = self.data_and_evaluations_from_raw_data(
raw_data=raw_data_by_arm,
metric_names=list(self.metric_names),
trial_index=trial.index,
sample_sizes=sample_sizes or {},
start_time=metadata.get("start_time"),
end_time=metadata.get("end_time"),
)
return evaluations, data
# ------------------------------ Validators. -------------------------------
@staticmethod
def _validate_can_complete_trial(trial: BaseTrial) -> None:
if trial.status.is_completed:
raise ValueError(
f"Trial {trial.index} has already been completed with data."
"To add more data to it (for example, for a different metric), "
"use `ax_client.update_trial_data`."
)
if trial.status.is_abandoned or trial.status.is_failed:
raise ValueError(
f"Trial {trial.index} has been marked {trial.status.name}, so it "
"no longer expects data."
)
def _validate_search_space_membership(self, parameters: TParameterization) -> None:
self.experiment.search_space.check_membership(
parameterization=parameters, raise_error=True
)
# `check_membership` uses int and float interchangeably, which we don't
# want here.
for p_name, parameter in self.experiment.search_space.parameters.items():
if (
isinstance(self.experiment.search_space, HierarchicalSearchSpace)
and p_name not in parameters
):
# Parameterizations in HSS-s can be missing some of the dependent
# parameters based on the hierarchical structure and values of
# the parameters those depend on.
continue
param_val = parameters.get(p_name)
if not isinstance(param_val, parameter.python_type):
typ = type(param_val)
raise ValueError(
f"Value for parameter {p_name}: {param_val} is of type {typ}, "
f"expected {parameter.python_type}. If the intention was to have"
f" the parameter on experiment be of type {typ}, set `value_type`"
f" on experiment creation for {p_name}."
)
def _validate_trial_data(self, trial: Trial, data: Data) -> None:
for metric_name in data.df["metric_name"].values:
if metric_name not in self.experiment.metrics:
logger.info(
f"Data was logged for metric {metric_name} that was not yet "
"tracked on the experiment. Please specify `tracking_metric_"
"names` argument in AxClient.create_experiment to add tracking "
"metrics to the experiment. Without those, all data users "
"specify is still attached to the experiment, but will not be "
"fetched in `experiment.fetch_data()`, but you can still use "
"`experiment.lookup_data_for_trial` to get all attached data."
)
def _validate_early_stopping_strategy(
self, support_intermediate_data: bool
) -> None:
if self._early_stopping_strategy is not None and not support_intermediate_data:
raise ValueError(
"Early stopping is only supported for experiments which allow "
" reporting intermediate trial data by setting passing "
"`support_intermediate_data=True`."
)
def __repr__(self) -> str:
"""String representation of this client."""
return f"{self.__class__.__name__}(experiment={self._experiment})"
# -------- Backward-compatibility with old save / load method names. -------
[docs] @staticmethod
def get_recommended_max_parallelism() -> None:
raise NotImplementedError(
"Use `get_max_parallelism` instead; parallelism levels are now "
"enforced in generation strategy, so max parallelism is no longer "
"just recommended."
)
[docs] @staticmethod
def load_experiment(experiment_name: str) -> None:
raise NotImplementedError(
"Use `load_experiment_from_database` to load from SQL database or "
"`load_from_json_file` to load optimization state from .json file."
)
[docs] @staticmethod
def load(filepath: Optional[str] = None) -> None:
raise NotImplementedError(
"Use `load_experiment_from_database` to load from SQL database or "
"`load_from_json_file` to load optimization state from .json file."
)
[docs] @staticmethod
def save(filepath: Optional[str] = None) -> None:
raise NotImplementedError(
"Use `save_to_json_file` to save optimization state to .json file."
)
