semsynth.downstream_fidelity

Downstream fidelity comparison between real and synthetic data.

Functions

auto_formula(df, meta, cfg)
compute_downstream(df_real, df_synth[, ...])
dataclass([cls, init, repr, eq, order, ...])	Add dunder methods based on the fields defined in the class.
get_column_name(entry, *[, extra_keys])	Return the first non-empty column name found in a JSON-LD column entry.
modeling_role(node, *[, default])	Return a simplified role label for modeling contexts.
recompute_downstream([metrics, real, synth, ...])	Recompute downstream metrics from existing real/synthetic CSVs.
rule(*[, name, phony, base_iri, prov_dir, ...])	Decorate a function as a build rule with automatic provenance.

Classes

Any(*args, **kwargs)	Special type indicating an unconstrained type.
CategoricalDtype([categories, ordered])	Type for categorical data with the categories and orderedness.
Config()	Base configuration container with TOML application helpers.
DownstreamConfig([m, burnin, ...])
InPath(*paths)	Marker for input paths where "-" maps to stdin.
LogisticRegression([penalty, C, l1_ratio, ...])	Logistic Regression (aka logit, MaxEnt) classifier.
OutPath(*paths)	Marker for output paths where "-" maps to stdout.
Path(*args, **kwargs)	PurePath subclass that can make system calls.
SimpleImputer(*[, missing_values, strategy, ...])	Univariate imputer for completing missing values with simple strategies.

class semsynth.downstream_fidelity.DownstreamConfig(m: 'int' = 20, burnin: 'int' = 5, max_interactions: 'int' = 5, cv: 'int' = 5)

Bases: Config

burnin: int = 5

cv: int = 5

m: int = 20

max_interactions: int = 5

semsynth.downstream_fidelity.auto_formula(df: DataFrame, meta: Mapping[str, Any], cfg: DownstreamConfig) → str

semsynth.downstream_fidelity.compute_downstream(df_real: DataFrame, df_synth: DataFrame, meta: Mapping[str, Any] | None = None, *, cfg: DownstreamConfig | None = None, target: str | None = None) → Mapping[str, Any]

semsynth.downstream_fidelity.recompute_downstream(metrics: OutPath = OutPath('{metrics}'), *, real: InPath | None = None, synth: InPath | None = None, meta: InPath | None = None, target: str | None = None, verbose: bool = False) → None

Recompute downstream metrics from existing real/synthetic CSVs.