neuraxle.metaopt.optimizer¶

Module-level documentation for neuraxle.metaopt.optimizer. Here is an inheritance diagram, including dependencies to other base modules of Neuraxle:

Neuraxle’s Hyperparameter Optimizer Base Classes¶

Not all hyperparameter optimizers are there, but the base can be found here.

See also

TreeParzenEstimator,

Classes

`BaseHyperparameterOptimizer`
`GridExplorationSampler`(expected_n_trials, seed_i)	This hyperparameter space optimizer is similar to a grid search, however, it does try to greedily sample maximally different points in the space to explore it.
`HyperparameterSamplerStub`(…)
`RandomSearchSampler`()	AutoML Hyperparameter Optimizer that randomly samples the space of random variables.

class neuraxle.metaopt.optimizer.BaseHyperparameterOptimizer[source]¶

Bases: abc.ABC

find_next_best_hyperparams(round: neuraxle.metaopt.data.reporting.RoundReport, hp_space: neuraxle.hyperparams.space.HyperparameterSpace) → neuraxle.hyperparams.space.HyperparameterSamples[source]¶

Find the next best hyperparams using previous trials, that is the whole neuraxle.metaopt.data.aggregate.Round.

Return type: HyperparameterSamples
Parameters: round (RoundReport) – a neuraxle.metaopt.data.aggregate.Round
Returns: next hyperparameter samples to train on

_abc_impl = <_abc_data object>¶

class neuraxle.metaopt.optimizer.HyperparameterSamplerStub(preconfigured_hp_samples: neuraxle.hyperparams.space.HyperparameterSamples)[source]¶

Bases: neuraxle.metaopt.optimizer.BaseHyperparameterOptimizer

__init__(preconfigured_hp_samples: neuraxle.hyperparams.space.HyperparameterSamples)[source]¶: Initialize self. See help(type(self)) for accurate signature.

find_next_best_hyperparams(round: neuraxle.metaopt.data.reporting.RoundReport, hp_space: neuraxle.hyperparams.space.HyperparameterSpace) → neuraxle.hyperparams.space.HyperparameterSamples[source]¶

Find the next best hyperparams using previous trials, that is the whole neuraxle.metaopt.data.aggregate.Round.

Return type: HyperparameterSamples
Parameters: round (RoundReport) – a neuraxle.metaopt.data.aggregate.Round
Returns: next hyperparameter samples to train on

_abc_impl = <_abc_data object>¶

class neuraxle.metaopt.optimizer.RandomSearchSampler[source]¶

Bases: neuraxle.metaopt.optimizer.BaseHyperparameterOptimizer

AutoML Hyperparameter Optimizer that randomly samples the space of random variables. Please refer to AutoML for a usage example.

See also

Trainer, HyperparamsRepository,

__init__()[source]¶: Initialize self. See help(type(self)) for accurate signature.

find_next_best_hyperparams(round: neuraxle.metaopt.data.reporting.RoundReport, hp_space: neuraxle.hyperparams.space.HyperparameterSpace) → neuraxle.hyperparams.space.HyperparameterSamples[source]¶

Randomly sample the next hyperparams to try.

Return type: HyperparameterSamples
Parameters: round (RoundReport) – round report
Returns: next random hyperparams

_abc_impl = <_abc_data object>¶

class neuraxle.metaopt.optimizer.GridExplorationSampler(expected_n_trials: int = 0, seed_i: int = 0)[source]¶

Bases: neuraxle.metaopt.optimizer.BaseHyperparameterOptimizer

This hyperparameter space optimizer is similar to a grid search, however, it does try to greedily sample maximally different points in the space to explore it. This space optimizer has a fixed pseudorandom exploration method that makes the sampling reproductible.

When over the expected_n_trials (if sampling too much), the sampler will turn to a non-seeded random search.

It may be good for space exploration before a TPE or for unit tests.

If the expected_n_trials is not set or set to 0, the sampler will guess its ideal sampling count and then switch to random search after that.

__init__(expected_n_trials: int = 0, seed_i: int = 0)[source]¶: Initialize self. See help(type(self)) for accurate signature.

static estimate_ideal_n_trials(hp_space: neuraxle.hyperparams.space.HyperparameterSpace) → int[source]¶

_reinitialize_grid(hp_space: neuraxle.hyperparams.space.HyperparameterSpace, previous_trials_hp: List[neuraxle.hyperparams.space.HyperparameterSamples]) → neuraxle.hyperparams.space.HyperparameterSamples[source]¶

Update the grid exploration sampler.

Return type: HyperparameterSamples
Parameters: round_scope – round scope
Returns: next random hyperparams

find_next_best_hyperparams(round: neuraxle.metaopt.data.reporting.RoundReport, hp_space: neuraxle.hyperparams.space.HyperparameterSpace) → neuraxle.hyperparams.space.HyperparameterSamples[source]¶

Sample the next hyperparams to try.

Return type: HyperparameterSamples
Parameters: round_scope – round scope
Returns: next hyperparams

_generate_grid(hp_space: neuraxle.hyperparams.space.HyperparameterSpace)[source]¶

Generate the grid of hyperparameters to pick from.

Parameters: hp_space (HyperparameterSpace) – hyperparameter space

static _pseudo_shuffle_list(x: List[T], seed: int = 0) → list[source]¶: Shuffle a list to create a pseudo-random order that is interesting.

_gen_keys_for_grid() → List[int][source]¶

Generate the keys for the grid.

Parameters: i – index
Returns: keys

_reshuffle_grid(new_sample: OrderedDict[str, Any] = None)[source]¶: Reshuffling with pseudo-random seed the hyperparameters’ values:

_abc_impl = <_abc_data object>¶