scitex_ml.classification.reporters.reporter_utils.data_models

Data models for classification reporting using dataclasses.

Provides type-safe, validated data structures for metrics and reports.

Classes

`ExperimentConfig`(name, output_dir[, ...])	Configuration for an experiment.
`ExperimentResults`(config[, fold_results, ...])	Complete results for an experiment.
`FoldResults`(fold_id[, balanced_accuracy, ...])	Results for a single fold.
`MetricResult`(metric_name, value[, fold, ...])	Single metric result with metadata.

class scitex_ml.classification.reporters.reporter_utils.data_models.MetricResult(metric_name, value, fold=None, timestamp=<factory>, metadata=<factory>)[source]

Single metric result with metadata.

metric_name: str

value: float | ndarray | DataFrame

fold: int | None = None

timestamp: datetime

metadata: Dict[str, Any]

to_dict()[source]

Convert to dictionary for JSON serialization.

Return type:: Dict[str, Any]

to_json()[source]

Convert to JSON string.

Return type:: str

save(path)[source]

Save metric to file based on type.

Return type:: None

__init__(metric_name, value, fold=None, timestamp=<factory>, metadata=<factory>)

class scitex_ml.classification.reporters.reporter_utils.data_models.FoldResults(fold_id, balanced_accuracy=None, mcc=None, confusion_matrix=None, classification_report=None, roc_auc=None, pr_auc=None, custom_metrics=<factory>)[source]

Results for a single fold.

fold_id: int

balanced_accuracy: MetricResult | None = None

mcc: MetricResult | None = None

confusion_matrix: MetricResult | None = None

classification_report: MetricResult | None = None

roc_auc: MetricResult | None = None

pr_auc: MetricResult | None = None

custom_metrics: Dict[str, MetricResult]

is_complete(required_metrics)[source]

Check if all required metrics are present.

Return type:: bool

get_available_metrics()[source]

Get list of available metrics.

Return type:: List[str]

to_dict()[source]

Convert to dictionary.

Return type:: Dict[str, Any]

__init__(fold_id, balanced_accuracy=None, mcc=None, confusion_matrix=None, classification_report=None, roc_auc=None, pr_auc=None, custom_metrics=<factory>)

class scitex_ml.classification.reporters.reporter_utils.data_models.ExperimentConfig(name, output_dir, n_folds=5, required_metrics=<factory>, classifier='unknown', dataset='unknown', features=<factory>, parameters=<factory>, random_seed=42, timestamp=<factory>)[source]

Configuration for an experiment.

name: str

output_dir: Path

n_folds: int = 5

required_metrics: List[str]

classifier: str = 'unknown'

dataset: str = 'unknown'

features: List[str]

parameters: Dict[str, Any]

random_seed: int = 42

timestamp: datetime

to_dict()[source]

Convert to dictionary.

Return type:: Dict[str, Any]

save()[source]

Save configuration to output directory.

Return type:: None

__init__(name, output_dir, n_folds=5, required_metrics=<factory>, classifier='unknown', dataset='unknown', features=<factory>, parameters=<factory>, random_seed=42, timestamp=<factory>)

class scitex_ml.classification.reporters.reporter_utils.data_models.ExperimentResults(config, fold_results=<factory>, summary_statistics=None)[source]

Complete results for an experiment.

config: ExperimentConfig

fold_results: List[FoldResults]

summary_statistics: Dict[str, Any] | None = None

add_fold(fold_result)[source]

Add results for a fold.

Return type:: None

validate_completeness()[source]

Validate that all required metrics are present for all folds.

Return type:: Dict[str, Any]

calculate_summary_statistics()[source]

Calculate summary statistics across folds.

Return type:: Dict[str, Any]

to_dataframe()[source]

Convert results to a DataFrame for easy analysis.

Return type:: DataFrame

save()[source]

Save all results to the output directory.

Return type:: None

__init__(config, fold_results=<factory>, summary_statistics=None)