bayesline.api.equity.RootRiskDatasetSettings#

model RootRiskDatasetSettings#

Settings for creating a root risk dataset directly from uploads.

A root dataset has no reference dataset. It is assembled by reading from seven upload sources (market_cap, price, exposures, exchange_dates, exchange_rates, series, idmap) and producing a ConsolidatedDataParams that can subsequently be referenced by the standard DerivedRiskDatasetSettings flow.

All upload rows are expected to be in master-id space (asset_id_type column equal to master_id_type). Foreign-id rows in the same uploads are tolerated but ignored during root-dataset construction.

Shares the joint RiskDatasetSettingsMenu with DerivedRiskDatasetSettings; the validator populates the available_*_sources fields relevant to this variant. The kind discriminator routes Pydantic-union dispatch.

Fields#
Field	Type	Required	Default
`as_of_date`	`str` \| `date` \| `datetime`	Yes
`dense_factor_groups`	`dict`[`str`, `str` \| `None`]	No	factory
`exchange_dates_source`	`str`	Yes
`exchange_rates_source`	`str`	Yes
`exposures_source`	`str`	Yes
`idmap_source`	`str`	Yes
`kind`	`Literal`[‘root’]	No	`'root'`
`market_cap_source`	`str`	Yes
`master_id_core_type`	`str`	No	`'bayesid_core'`
`master_id_type`	`str`	No	`'bayesid'`
`price_source`	`str`	Yes
`series_source`	`str` \| `None`	No	`None`
`sparse_factor_groups`	`dict`[`str`, `str` \| `None`]	No	factory

field as_of_date: str | date | datetime [Required]#: Snapshot date for the IdMapper. Only id-mapping rows whose start_date <= as_of_date are retained when constructing the dataset’s IdMapper. Pick the date as of which the uploaded idmap should be considered authoritative — typically the most recent date for which the upload contains complete mappings. Accepts a date, datetime, or ISO-format string; the consumer normalizes via to_date(...).

field dense_factor_groups: dict[str, str | None] [Optional]#: Mapping of dense factor-group name to hierarchy upload source. Use None as the value to use a default flat hierarchy.

field exchange_dates_source: str [Required]#: Upload source name providing trading-day membership per exchange: (date, exchange).

field exchange_rates_source: str [Required]#: Upload source name providing single-base USD FX rates: (date, ccy, fx_rate).

field exposures_source: str [Required]#: Upload source name providing factor exposures: (date, asset_id, asset_id_type, factor_group, factor, exposure).

field idmap_source: str [Required]#: Upload source name providing the id-map table: (start_date, end_date, from_id_type, from_id, to_id_type, to_id). The (master_id_type -> master_id_core_type) projection is extracted during dataset construction.

field kind: Literal['root'] [Optional]#: Discriminator field for the risk dataset settings union. root selects this variant for datasets built directly from upload sources (no reference dataset).

field market_cap_source: str [Required]#: Upload source name providing per-asset-day market-cap data: (date, asset_id, asset_id_type, ccy, market_cap, volume, idio_vol).

field master_id_core_type: str [Optional]#: The canonical core-asset-id type. Used to extract the asset-to-core mapping from the idmap upload.

field master_id_type: str [Optional]#: The canonical asset-id type. Upload rows whose asset_id_type does not equal this value are ignored.

field price_source: str [Required]#: Upload source name providing prices: (date, asset_id, asset_id_type, ccy, close, return, delisted).

field series_source: str | None [Optional]#: Upload source name providing time series in long format: (date, series, value). Optional.

field sparse_factor_groups: dict[str, str | None] [Optional]#: Mapping of sparse factor-group name to hierarchy upload source. Use None as the value to use a default flat hierarchy.