synthcity.metrics.eval_detection module

class DetectionEvaluator(**kwargs: Any)

Bases: synthcity.metrics.core.metric.MetricEvaluator

Inheritance diagram of synthcity.metrics.eval_detection.DetectionEvaluator

Train a SKLearn classifier to detect the synthetic data from real data.

Synthetic and real data are combined to form a new dataset. K-fold cross validation is performed to see how well a classifier can distinguish real from synthetic.

Returns

The average AUCROC score for detecting synthetic data.

Score:

0: The datasets are indistinguishable. 1: The datasets are totally distinguishable.

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 SyntheticDetectionGMM(**kwargs: Any)

Bases: synthcity.metrics.eval_detection.DetectionEvaluator

Inheritance diagram of synthcity.metrics.eval_detection.SyntheticDetectionGMM

Train a GaussianMixture model to detect synthetic data.

Returns

The average score for detecting synthetic data.

Score:

0: The datasets are indistinguishable. 1: The datasets are totally distinguishable.

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 SyntheticDetectionLinear(**kwargs: Any)

Bases: synthcity.metrics.eval_detection.DetectionEvaluator

Inheritance diagram of synthcity.metrics.eval_detection.SyntheticDetectionLinear

Train a LogisticRegression classifier to detect the synthetic data.

Returns

The average AUCROC score for detecting synthetic data.

Score:

0: The datasets are indistinguishable. 1: The datasets are totally distinguishable.

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 SyntheticDetectionMLP(**kwargs: Any)

Bases: synthcity.metrics.eval_detection.DetectionEvaluator

Inheritance diagram of synthcity.metrics.eval_detection.SyntheticDetectionMLP

Train a MLP classifier to detect the synthetic data.

Returns

The average AUCROC score for detecting synthetic data.

Score:

0: The datasets are indistinguishable. 1: The datasets are totally distinguishable.

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 SyntheticDetectionXGB(**kwargs: Any)

Bases: synthcity.metrics.eval_detection.DetectionEvaluator

Inheritance diagram of synthcity.metrics.eval_detection.SyntheticDetectionXGB

Train a XGBoostclassifier to detect the synthetic data.

Returns

The average AUCROC score for detecting synthetic data.

Score:

0: The datasets are indistinguishable. 1: The datasets are totally distinguishable.

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