Configuration_Setup.md

graph LR
    Settings["Settings"]
    SettingsCreator["SettingsCreator"]
    JSONSchemaValidator["JSONSchemaValidator"]
    SettingsColumnCleaner["SettingsColumnCleaner"]
    InvalidColumnsLogger["InvalidColumnsLogger"]
    SettingsCreator -- "creates" --> Settings
    JSONSchemaValidator -- "validates" --> Settings
    SettingsColumnCleaner -- "processes" --> Settings
    InvalidColumnsLogger -- "validates" --> Settings
    InvalidColumnsLogger -- "uses" --> SettingsColumnCleaner

Loading

Details

The Settings object serves as the core configuration for the Splink linkage process. The SettingsCreator initiates this process by constructing the Settings object from various inputs. Once created, the Settings object undergoes validation by the JSONSchemaValidator to ensure its structural and data type integrity against a predefined schema. Subsequently, the SettingsColumnCleaner refines the Settings by validating and cleaning column names against actual input dataframes, ensuring data schema consistency. Finally, the InvalidColumnsLogger identifies and reports any remaining issues with column references within SQL expressions defined in the Settings, often leveraging the cleaned column information from SettingsColumnCleaner to provide detailed feedback for debugging. This sequential flow ensures a robust and validated configuration for the linkage operations.

Settings

Serves as the central, immutable data model that encapsulates all linkage configuration, including blocking rules, comparison levels, and model parameters. It acts as the single source of truth for the entire linkage process. Its fundamental importance lies in providing a standardized, consistent representation of the user's linkage intent.

Related Classes/Methods:

• splink.internals.settings.Settings:183-696

SettingsCreator

Responsible for parsing raw input (e.g., dictionaries, JSON files) and instantiating the Settings object. It prepares the initial configuration for subsequent validation steps. This component is crucial as it bridges the gap between user-defined input and the internal data model.

Related Classes/Methods:

• splink.internals.settings_creator.SettingsCreator:17-137

JSONSchemaValidator

Enforces the structural and data type correctness of the Settings object by validating it against a predefined JSON schema. This ensures adherence to expected formats and prevents malformed configurations, acting as the first line of defense for configuration integrity.

Related Classes/Methods:

• splink.internals.validate_jsonschema.JSONSchemaValidator:1-100

SettingsColumnCleaner

Focuses specifically on validating and cleaning column names referenced within the Settings object (e.g., unique IDs, columns used in blocking rules and comparison levels) against the actual input dataframes. It ensures that all referenced columns exist and are correctly specified, which is vital for preventing runtime errors related to data schema mismatches.

Related Classes/Methods:

• splink.internals.settings_validation.settings_column_cleaner.SettingsColumnCleaner:119-155

InvalidColumnsLogger

Identifies and logs specific issues related to missing or invalid columns within SQL expressions defined in blocking rules and comparison levels. It provides detailed, actionable feedback on configuration errors, aiding in debugging. This component is essential for user experience and efficient problem resolution.

Related Classes/Methods:

• splink.internals.settings_validation.log_invalid_columns.InvalidColumnsLogger:181-253

FAQ