# Author: LKouadio <etanoyau@gmail.com>
# License: LGPL-3.0
"""Domain-gap and noise simulation for AI-assisted EM inversion (M3).
This package turns a clean :class:`~pycsamt.ai.data.contracts.SurveyData`
into realistic training data by injecting heteroscedastic noise, dropout,
static shift, galvanic distortion, coordinate perturbation, and outliers
(:mod:`~pycsamt.ai.domain_gap.simulator`); by fitting plausible parameter
ranges from a real survey's own QC diagnostics, AMT, CSAMT, MT, or
otherwise (:mod:`~pycsamt.ai.domain_gap.survey_fit`); and by comparing
simulated and field feature distributions quantitatively
(:mod:`~pycsamt.ai.domain_gap.report`).
"""
from __future__ import annotations
from .audit import (
DimensionalitySummary,
FrequencyGridReport,
StationExclusion,
SurveyAuditReport,
audit_survey,
)
from .empirical import (
EmpiricalCorruptionResult,
apply_empirical_corruption,
)
from .report import (
DistributionComparisonReport,
FeatureComparison,
compare_feature_distributions,
compare_survey_distributions,
)
from .simulator import (
SEVERITY_PRESETS,
CorruptionConfig,
CorruptionRecord,
add_heteroscedastic_noise,
apply_corruption_suite,
apply_dropout,
apply_error_floor,
apply_galvanic_distortion,
apply_static_shift,
inject_outliers,
perturb_coordinates,
)
from .survey_fit import (
fit_corruption_config,
fit_distortion_priors_from_sites,
survey_data_from_sites,
)
__all__ = [
"StationExclusion",
"FrequencyGridReport",
"DimensionalitySummary",
"SurveyAuditReport",
"audit_survey",
"EmpiricalCorruptionResult",
"apply_empirical_corruption",
"SEVERITY_PRESETS",
"CorruptionConfig",
"CorruptionRecord",
"add_heteroscedastic_noise",
"apply_corruption_suite",
"apply_dropout",
"apply_error_floor",
"apply_galvanic_distortion",
"apply_static_shift",
"inject_outliers",
"perturb_coordinates",
"fit_corruption_config",
"fit_distortion_priors_from_sites",
"survey_data_from_sites",
"DistributionComparisonReport",
"FeatureComparison",
"compare_feature_distributions",
"compare_survey_distributions",
]
