Environment

Run-level settings gathered in one place: data sources, date format, calendar and simulation defaults.

environment

Session configuration.

An :class:Environment groups settings into categories (data sources, data formats, calendar, simulation) and sets them by flat name through set_environment(**kwargs). Its main use is to tell beacon.data.loader.load_data where the market and reference data are, either as in-memory DataFrames or as file paths.

Environment

Environment()

Centralized configuration for a Beacon session.

Settings are grouped into category dataclasses and can be set via set_environment(**kwargs) using flat parameter names.

Attributes:

Name	Type	Description
data_source		Where the input data comes from.
data		How input data is parsed.
calendar		Calendar conventions.
simulation		Simulation settings.

set_environment

set_environment(**kwargs: Any) -> None

Set one or more parameters by their flat field names.

Names are checked against every category's fields before anything is set, so an unknown name changes nothing. Values are not checked.

Parameters:

Name	Type	Description	Default
**kwargs	Any	Field name to new value, e.g. MARKET_DATA_PATH="prices.csv".	{}

Raises:

Type	Description
ValueError	If any name is not a field of a settings category.

summary

summary() -> dict[str, dict[str, Any]]

Return all current settings as a nested dict.

Returns:

Name	Type	Description
dict	dict[str, dict[str, Any]]	Category name to {field name: value}. DataFrame values
	dict[str, dict[str, Any]]	are left out.

config

Environment configuration dataclasses and the central Environment class.

beacon.data.loader.load_data reads the market and reference data settings and DATE_FORMAT from an Environment. The other fields are stored, and reported by :meth:Environment.summary, but nothing else in Beacon reads them.

DataSourceConfig dataclass

DataSourceConfig(
    MARKET_DATA_PATH: str | None = None,
    REFERENCE_DATA_PATH: str | None = None,
    FUNDAMENTALS_DATA_PATH: str | None = None,
    CORPORATE_ACTIONS_DATA_PATH: str | None = None,
    FX_RATES_DATA_PATH: str | None = None,
    MARKET_DATA: DataFrame | None = None,
    REFERENCE_DATA: DataFrame | None = None,
    FUNDAMENTALS_DATA: DataFrame | None = None,
    CORPORATE_ACTIONS_DATA: DataFrame | None = None,
    FX_RATES_DATA: DataFrame | None = None,
)

Where the input data comes from.

Each dataset can be given either as an in-memory DataFrame or as a file path. When both are set for the same dataset, load_data uses the DataFrame.

Attributes:

Name	Type	Description
MARKET_DATA_PATH	str | None	File to read market data (prices) from.
REFERENCE_DATA_PATH	str | None	File to read reference data from.
FUNDAMENTALS_DATA_PATH	str | None	File to read fundamentals data from.
CORPORATE_ACTIONS_DATA_PATH	str | None	File to read corporate actions from.
FX_RATES_DATA_PATH	str | None	File to read FX rates from.
MARKET_DATA	DataFrame | None	Market data as a long-form DataFrame.
REFERENCE_DATA	DataFrame | None	Reference data as a DataFrame.
FUNDAMENTALS_DATA	DataFrame | None	Fundamentals data as a DataFrame.
CORPORATE_ACTIONS_DATA	DataFrame | None	Corporate actions as a DataFrame.
FX_RATES_DATA	DataFrame | None	FX rates as a DataFrame.

DataConfig dataclass

DataConfig(DATE_FORMAT: str = '%Y-%m-%d')

How input data is parsed.

Attributes:

Name	Type	Description
DATE_FORMAT	str	The strftime format of the market data's dates.

CalendarConfig dataclass

CalendarConfig(
    TRADING_DAYS: int = 252,
    DAY_COUNT_CONVENTION: str = "ACT/365",
)

Calendar conventions.

Attributes:

Name	Type	Description
TRADING_DAYS	int	Trading days per year.
DAY_COUNT_CONVENTION	str	Day-count convention name, such as "ACT/365".

SimulationConfig dataclass

SimulationConfig(
    TRANSACTION_COST: float = 0.0,
    FLOAT_TOLERANCE: float = 1e-09,
)

Simulation settings.

Attributes:

Name	Type	Description
TRANSACTION_COST	float	Transaction cost.
FLOAT_TOLERANCE	float	Tolerance for floating-point comparisons.

IndexConfig dataclass

IndexConfig()

Reserved for index settings. It holds none and is not part of Environment.

Environment

Environment()

Centralized configuration for a Beacon session.

Settings are grouped into category dataclasses and can be set via set_environment(**kwargs) using flat parameter names.

Attributes:

Name	Type	Description
data_source		Where the input data comes from.
data		How input data is parsed.
calendar		Calendar conventions.
simulation		Simulation settings.

set_environment

set_environment(**kwargs: Any) -> None

Set one or more parameters by their flat field names.

Names are checked against every category's fields before anything is set, so an unknown name changes nothing. Values are not checked.

Parameters:

Name	Type	Description	Default
**kwargs	Any	Field name to new value, e.g. MARKET_DATA_PATH="prices.csv".	{}

Raises:

Type	Description
ValueError	If any name is not a field of a settings category.

summary

summary() -> dict[str, dict[str, Any]]

Return all current settings as a nested dict.

Returns:

Name	Type	Description
dict	dict[str, dict[str, Any]]	Category name to {field name: value}. DataFrame values
	dict[str, dict[str, Any]]	are left out.