Analyzer

systematica.tuners.sqlite.analyzer

SQLiteOptunaAnalyzer

SQLiteOptunaAnalyzer(
    project: str,
    api_token: str = None,
    init_kwargs: Dict[str, Any] = _Nothing.NOTHING,
)

SQLite-based optimization result analyzer.

Queries Optuna’s native SQLite tables to fetch and analyze optimization results. No custom tables - all data retrieved directly from Optuna’s schema.

Method generated by attrs for class SQLiteOptunaAnalyzer.

Ancestors

• systematica.tuners.base.BaseOptunaAnalyzer
• systematica.tuners.base.BaseAnalyzer
• abc.ABC

Static methods

from_storage_url

from_storage_url(
    storage_url: str = 'sqlite:////Users/jmr/GitHub/systematica/sma_sqlite_optuna.db',
    **init_kwargs: Dict[str, Any],
) ‑> systematica.tuners.sqlite.analyzer.SQLiteOptunaAnalyzer

Create analyzer instance from storage URL.

Parameters:

Name	Type	Default	Description
storage_url	str	DEFAULT_STORAGE_URL	SQLite storage URL.
init_kwargs	tp.Kwargs	--	Additional initialization kwargs.

Returns:

Type	Description
SQLiteOptunaAnalyzer	Initialized analyzer instance.

Instance variables

• init_kwargs: Dict[str, Any]: Initialization kwargs for storage configuration.

• storage_url: str: SQLite storage URL.

Methods

run_context

run_context(
    self,
    **kwargs,
)

Context manager for consistency with base class API.

Parameters:

Name	Type	Default	Description
kwargs	tp.Kwargs	--	Additional arguments (unused).

Yields:

Type	Description
None	No session needed - Optuna handles database connections internally.

fetch_trials

fetch_trials(
    self,
    metadata_tag: str,
    run_ids: Tuple[str],
    is_trials: bool,
    is_file: bool = False,
    mode: str = 'read-only',
) ‑> pandas.core.frame.DataFrame

Fetch optimization data from Optuna’s native tables.

Parameters:

Name	Type	Default	Description
metadata_tag	str	--	Data type to fetch: - ‘best/params’: Best trial parameters - ‘best/values’: Best trial metric values - ‘study/distributions’: Parameter distributions - ‘trials/trials’: All trial data
run_ids	tuple	--	of str Study names to fetch.
is_trials	bool	--	Whether fetching trial-level data (True) or study-level data (False).
is_file	bool	False	Unused, for API compatibility.
mode	str	"read-only"	Unused, for API compatibility with base class.

Returns:

Type	Description
pd.DataFrame	DataFrame with fetched data.

fetch_feature_config

fetch_feature_config(
    self,
    session,
    run_id: str,
) ‑> Dict[str, Any]

Fetch feature_config from Optuna’s study_user_attributes.

Parameters:

Name	Type	Default	Description
session	Any	--	Unused, kept for API compatibility with base class.
run_id	str	--	Study name.

Returns:

Type	Description
dict	Feature configuration dictionary.

Raises:

Type	Description
SystematicaError	If feature_config not found in study attributes.

fetch_optuna_study

fetch_optuna_study(
    self,
    run_id: str,
) ‑> optuna.study.study.Study

Load Optuna study from SQLite storage.

Parameters:

Name	Type	Default	Description
run_id	str	--	Study name.

Returns:

Type	Description
optuna.Study	Loaded Optuna study object with all user attributes intact.