smac.runhistory.encoder.boing_encoder

Classes

RunHistoryRawEncoder(scenario[, ...])

A transformer that transform the RunHistroy to vectors.

RunHistoryRawScaledEncoder(scenario[, ...])

Interfaces

class smac.runhistory.encoder.boing_encoder.RunHistoryRawEncoder(scenario, considered_states=[<StatusType.SUCCESS: 1>, <StatusType.CRASHED: 3>, <StatusType.MEMORYOUT: 5>, <StatusType.DONOTADVANCE: 6>], lower_budget_states=[], scale_percentage=5, seed=None)[source]

Bases: RunHistoryEncoder

A transformer that transform the RunHistroy to vectors. This set of classes will return the raw cost values in addition to the transformed cost values. The raw cost values can then be applied for local BO approaches.

transform_raw_values(values)[source]

Returns the raw input values before transformation.

Return type:

ndarray

transform_response_values(values)[source]

Returns the input values.

Return type:

ndarray

transform_with_raw(runhistory, budget_subset=None)[source]

Returns vector representation of runhistory; if imputation is disabled, censored (TIMEOUT with time < cutoff) will be skipped. This function returns both the raw and transformed cost values

Parameters:
Return type:

tuple[ndarray, ndarray, ndarray]

Returns:

  • X (numpy.ndarray) – configuration vector x instance features

  • Y (numpy.ndarray) – cost values

  • Y_raw (numpy.ndarray) – cost values before transformation

class smac.runhistory.encoder.boing_encoder.RunHistoryRawScaledEncoder(scenario, considered_states=[<StatusType.SUCCESS: 1>, <StatusType.CRASHED: 3>, <StatusType.MEMORYOUT: 5>, <StatusType.DONOTADVANCE: 6>], lower_budget_states=[], scale_percentage=5, seed=None)[source]

Bases: RunHistoryRawEncoder, RunHistoryLogScaledEncoder

transform_raw_values(values)[source]

Returns the raw input values before transformation.

Return type:

ndarray

previous

smac.runhistory.encoder.abstract_encoder

next

smac.runhistory.encoder.eips_encoder