marketflows.config

Configuration loading and validation.

Parses TOML settings into dataclasses, applies defaults, and enforces basic sanity/consistency checks for providers, analysis, and plotting.

Functions

`get_provider_credentials`(provider[, ...])
`load_and_validate_config`(config_file)	Load and validate the config file.

Classes

`AnalysisConfig`(provider_config, ...[, ...])	Create a dataclass to hold the configuration values for analysis.
`PlotConfig`(hours_ago)	Create a dataclass to hold the configuration values for plots.
`ProviderConfig`(provider, days, flow_types[, ...])	Create a dataclass to hold the configuration values for the providers.

class marketflows.config.AnalysisConfig(provider_config, smoothing_ema, is_unit_normalize, diff_orders=<factory>, ema_periods=<factory>)

Bases: object

Create a dataclass to hold the configuration values for analysis.

provider_config: provider config

smoothing_ema: EMA period for smoothing growth and inflection data

is_unit_normalize: will we be creating plots that are normalized at each time step?

diff_orders

the orders of differentiation where 0 is no differentiation, 1 is the first derivative (growth), and 2 is the 2nd derivative (inflection)

Type:: list[int]

ema_periods: EMA periods

class marketflows.config.PlotConfig(hours_ago)

Bases: object

Create a dataclass to hold the configuration values for plots.

hours_ago: list of offsets for tables columns

class marketflows.config.ProviderConfig(provider, days, flow_types, base_assets=<factory>, narratives=<factory>, range_lower_limits=<factory>, asset_groups=<factory>)

Bases: object

Create a dataclass to hold the configuration values for the providers.

provider

name of provider. Leave as “” to read saved data from last run

Type:: str

days

the number of days we want to collect data for

Type:: int

flow_types

the list of flow types we want to collect data for

Type:: list[FlowType]

base_assets

assets used as base currency (e.g. EUR, JPY, etc.). These must be in decreasing MC order.

Type:: list[str]

narratives

list of narratives to be graphed

Type:: list[str]

range_lower_limits

the lower limits of each of the market cap ranges

Type:: list[float]

asset_groups

groups of assets as would be the case for multiple portfolios containing (often) different assets

Type:: dict[str, set[str]]

Examples

>>> ProviderConfig(provider="td-ameritrade", days=10, flow_types=["narratives"], narratives=["ai"])
ProviderConfig(provider='td-ameritrade', days=10, flow_types=['narratives'], base_assets=['us-dollar'], narratives=['ai'], range_lower_limits=[], asset_groups={})

marketflows.config.load_and_validate_config(config_file)

Load and validate the config file.

This does deep merge (i.e. it will descend into each nested item and only replace matching keys.

Return type:: tuple[ProviderConfig, AnalysisConfig, PlotConfig]