mlos_bench.storage.base_trial_data

Base interface for accessing the stored benchmark trial data.

A single trial is a single run of an experiment with a given configuration (e.g., set of tunable parameters).

Classes

TrialData

Base interface for accessing the stored experiment benchmark trial data.

Module Contents

class mlos_bench.storage.base_trial_data.TrialData(*, experiment_id: str, trial_id: int, tunable_config_id: int, ts_start: datetime.datetime, ts_end: datetime.datetime | None, status: mlos_bench.environments.status.Status)[source]

Base interface for accessing the stored experiment benchmark trial data.

A trial is a single run of an experiment with a given configuration (e.g., set of tunable parameters).

Parameters:
__eq__(other: Any) bool[source]
Parameters:

other (Any)

Return type:

bool

__repr__() str[source]
Return type:

str

property experiment_id: str[source]

ID of the experiment this trial belongs to.

Return type:

str

property metadata_df: pandas.DataFrame[source]
Abstractmethod:

Return type:

pandas.DataFrame

Retrieve the trials’ metadata parameters as a dataframe.

Note: this corresponds to the Trial object’s “config” property.

Returns:

metadata – An optional dataframe with the metadata associated with the trial. It has two str columns, “parameter” and “value”. Returns an empty dataframe if there is no metadata.

Return type:

pandas.DataFrame

property metadata_dict: dict[source]

Retrieve the trials’ metadata parameters as a dict.

Note: this corresponds to the Trial object’s “config” property.

Returns:

metadata

Return type:

dict

property results_df: pandas.DataFrame[source]
Abstractmethod:

Return type:

pandas.DataFrame

Retrieve the trials’ results from the storage.

Returns:

results – A dataframe with the trial results. It has two str columns, “metric” and “value”. If the trial status is not SUCCEEDED, the dataframe is empty.

Return type:

pandas.DataFrame

property results_dict: Dict[str, mlos_bench.tunables.tunable.TunableValue | None][source]

Retrieve the trials’ results from the storage as a dict.

Returns:

results

Return type:

dict

property status: mlos_bench.environments.status.Status[source]

Status of the trial.

Return type:

mlos_bench.environments.status.Status

property telemetry_df: pandas.DataFrame[source]
Abstractmethod:

Return type:

pandas.DataFrame

Retrieve the trials’ telemetry from the storage as a dataframe.

Returns:

config – A dataframe with the trial telemetry, if there is any. It has one datetime column, “ts”, and two str columns, “metric” and “value”. If the trial status is not SUCCEEDED, or there is no telemetry data, the dataframe is empty.

Return type:

pandas.DataFrame

property trial_id: int[source]

ID of the trial.

Return type:

int

property ts_end: datetime.datetime | None[source]

End timestamp of the trial (UTC).

Return type:

Optional[datetime.datetime]

property ts_start: datetime.datetime[source]

Start timestamp of the trial (UTC).

Return type:

datetime.datetime

property tunable_config: mlos_bench.storage.base_tunable_config_data.TunableConfigData[source]
Abstractmethod:

Return type:

mlos_bench.storage.base_tunable_config_data.TunableConfigData

Retrieve the trials’ tunable configuration data from the storage.

Note: this corresponds to the Trial object’s “tunables” property.

Returns:

tunable_config – A TunableConfigData object.

Return type:

TunableConfigData

property tunable_config_id: int[source]

ID of the (tunable) configuration of the trial.

Return type:

int

property tunable_config_trial_group: mlos_bench.storage.base_tunable_config_trial_group_data.TunableConfigTrialGroupData[source]
Abstractmethod:

Return type:

mlos_bench.storage.base_tunable_config_trial_group_data.TunableConfigTrialGroupData

Retrieve the trial’s (tunable) config trial group data from the storage.