synthcity.metrics.eval_sanity module

class BasicMetricEvaluator(**kwargs: Any)

Bases: synthcity.metrics.core.metric.MetricEvaluator

Inheritance diagram of synthcity.metrics.eval_sanity.BasicMetricEvaluator

static direction() → str

evaluate(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → float

evaluate_default(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → float

classmethod fqdn() → str

static name() → str

reduction() → Callable

static type() → str

use_cache(path: pathlib.Path) → bool

class CloseValuesProbability(**kwargs: Any)

Bases: synthcity.metrics.eval_sanity.BasicMetricEvaluator

Inheritance diagram of synthcity.metrics.eval_sanity.CloseValuesProbability

Compute the probability of close values between the real and synthetic data.

Score:: 0 means there is no chance to have synthetic rows similar to the real. 1 means that all the synthetic rows are similar to some real rows.

static direction() → str

evaluate(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → Dict

evaluate_default(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → float

classmethod fqdn() → str

static name() → str

reduction() → Callable

static type() → str

use_cache(path: pathlib.Path) → bool

class CommonRowsProportion(**kwargs: Any)

Bases: synthcity.metrics.eval_sanity.BasicMetricEvaluator

Inheritance diagram of synthcity.metrics.eval_sanity.CommonRowsProportion

Returns the proportion of rows in the real dataset leaked in the synthetic dataset.

Score:: 0: there are no common rows between the real and synthetic datasets. 1: all the rows in the real dataset are leaked in the synthetic dataset.

static direction() → str

evaluate(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → Dict

evaluate_default(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → float

classmethod fqdn() → str

static name() → str

reduction() → Callable

static type() → str

use_cache(path: pathlib.Path) → bool

class DataMismatchScore(**kwargs: Any)

Bases: synthcity.metrics.eval_sanity.BasicMetricEvaluator

Inheritance diagram of synthcity.metrics.eval_sanity.DataMismatchScore

Basic sanity score. Compares the data types between the column of the ground truth and the synthetic data.

Score:: 0: no datatype mismatch. 1: complete data type mismatch between the datasets.

static direction() → str

evaluate(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → Dict

evaluate_default(X_gt: synthcity.plugins.core.dataloader.DataLoader, X_syn: synthcity.plugins.core.dataloader.DataLoader) → float

classmethod fqdn() → str

static name() → str

reduction() → Callable

static type() → str

use_cache(path: pathlib.Path) → bool

class DistantValuesProbability(**kwargs: Any)

Bases: synthcity.metrics.eval_sanity.BasicMetricEvaluator

Inheritance diagram of synthcity.metrics.eval_sanity.DistantValuesProbability

Compute the probability of distant values between the real and synthetic data.

Score:: 0 means there is no chance to have rows in the synthetic far away from the real data. 1 means all the synthetic datapoints are far away from the real data.