autoemx.config.schema_models.quantification module

class autoemx.config.schema_models.quantification.QuantificationDiagnostics(*, iterations_run: int | None = None, converged: bool | None = None, interrupted: bool | None = None, min_background_ref_lines: float | None = None, missing_reference_peaks: List[str] | None = None, added_elements: List[str] | None = None, added_unquantified_elements: List[str] | None = None, possible_elements: List[str] | None = None, identification_note: str | None = None)

Bases: BaseModel

Execution diagnostics captured during iterative quantification.

Variables:

• iterations_run (int, optional) – Number of quantification iterations performed.

• converged (bool, optional) – True if the iterative fit converged within the allowed iterations.

• interrupted (bool, optional) – True when no composition result was produced, either because the spectrum failed a pre-fit validity check (insufficient counts, background too low, etc.) or because the iterative fit was aborted early after detecting signs of an unreliable result (poor fit quality, excessive analytical error, or excessive X-ray absorption) while interrupt_fits_bad_spectra=True. Spectra with interrupted=True are automatically re-quantified on the next run when interrupt_fits_bad_spectra=False.

• min_background_ref_lines (float, optional) – Minimum background counts under any reference peak used for quantification.

• missing_reference_peaks (list of str, optional) – Reference peaks that were absent or below the minimum acceptable PB ratio.

• added_elements (list of str, optional) – Elements found by the identification of missing elements, added to the fit and quantified.

• added_unquantified_elements (list of str, optional) – Elements found by the identification of missing elements and added to the fit, but not quantified (no standard).

• possible_elements (list of str, optional) – Elements whose presence is suggested by the identification of missing elements, but not confirmed by the fit (only reported, for manual checking).

• identification_note (str, optional) – Note of the identification of missing elements (e.g. unreliable background of the starting fit).

iterations_run: int | None

converged: bool | None

interrupted: bool | None

min_background_ref_lines: float | None

missing_reference_peaks: List[str] | None

added_elements: List[str] | None

added_unquantified_elements: List[str] | None

possible_elements: List[str] | None

identification_note: str | None

model_config = {'extra': 'forbid'}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

classmethod validate_iterations_run(v: int | None) → int | None

classmethod validate_min_background_ref_lines(value: float | None) → float | None

class autoemx.config.schema_models.quantification.QuantificationConfig(*, quantification_id: int, label: str | None = None, autoemx_version: str = '<=1.4.0', sample_elements: List[str] = <factory>, substrate_elements: List[str] = <factory>, els_w_fr: Dict[str, float] | None=None, options: Dict[str, ~typing.Any]=<factory>, reference_values_by_el_line: Dict[str, ~typing.Any]=<factory>, reference_lines_by_element: Dict[str, str]=<factory>, active_clustering_analysis_index: int | None = None, clustering_analyses: List[ClusteringAnalysis] = <factory>)

Bases: BaseModel

Configuration descriptor for one full quantification run.

The options mapping stores quantification-method options used to decide whether a previous run can be resumed (for example use_project_specific_std_dict).

The reference_lines_by_element mapping records which spectral line was selected for each element during this quantification run. This enables comparison between runs to detect whether line selection changed, which would require requantification.

NON_FINGERPRINT_OPTIONS: ClassVar[Tuple[str, ...]] = ('is_known_phase_mixture',)

quantification_id: int

label: str | None

autoemx_version: str

sample_elements: List[str]

substrate_elements: List[str]

els_w_fr: Dict[str, float] | None

options: Dict[str, Any]

reference_values_by_el_line: Dict[str, Any]

reference_lines_by_element: Dict[str, str]

active_clustering_analysis_index: int | None

clustering_analyses: List[ClusteringAnalysis]

model_config = {'extra': 'forbid'}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

classmethod validate_quantification_id_input(v: Any) → Any

classmethod validate_quantification_id(v: int) → int

classmethod validate_element_lists(values: List[str]) → List[str]

classmethod validate_els_w_fr(value: Dict[str, float] | None) → Dict[str, float] | None

classmethod move_options_min_bckgrnd_cnts(data: Any) → Any

Legacy compatibility: runs saved by AutoEMX 1.4.1 stored min_bckgrnd_cnts in their options.

It is a clustering setting (ClusteringConfig.min_bckgrnd_cnts): the value is moved to the clustering configs of the run that do not store it, so that their fingerprint is unchanged.

classmethod validate_options(value: Dict[str, Any]) → Dict[str, Any]

classmethod validate_reference_values_by_el_line(value: Dict[str, Any]) → Dict[str, Any]

classmethod validate_reference_lines_by_element(v: Dict[str, str]) → Dict[str, str]

Validate that element→line mapping contains no empty keys or values.

validate_clustering_analysis_index() → QuantificationConfig

Ensure active clustering analysis index is valid and defaults to the last entry.

get_active_clustering_analysis() → ClusteringAnalysis | None

Return the active clustering analysis, defaulting to the last one when available.

set_active_clustering_result(result: ClusteringResult) → None

Attach a clustering result to the active clustering analysis.

static derive_reference_lines_by_element(reference_values_by_el_line: Dict[str, Any], preferred_lines: Sequence[str] | None = None) → Dict[str, str]

Build a deterministic element->line map from available reference values.

fingerprint_payload() → Dict[str, Any]

Return canonical scientific inputs used to decide quantification reuse.

reference_values_by_el_line / reference_lines_by_element are scoped to sample_elements only (substrate reference lines are never part of the scientific signature and should not be stored).

fingerprint() → str

Compute deterministic SHA-256 fingerprint for quantification scientific inputs.

fingerprint_differences(other: QuantificationConfig) → Dict[str, Dict[str, Any]]

Return a deterministic diff of scientific inputs against another config.

class autoemx.config.schema_models.quantification.QuantificationResult(*, quantification_id: int, quant_flag: int | None = None, comment: str | None = None, composition_atomic_fractions: Dict[str, float] | None = None, composition_weight_fractions: Dict[str, float] | None = None, analytical_error: float | None = None, fit_result: FitResult | None = None, diagnostics: QuantificationDiagnostics | None = None)

Bases: BaseModel

Persisted per-spectrum quantification output for a specific run/config.

quantification_id: int

quant_flag: int | None

comment: str | None

composition_atomic_fractions: Dict[str, float] | None

composition_weight_fractions: Dict[str, float] | None

analytical_error: float | None

fit_result: FitResult | None

diagnostics: QuantificationDiagnostics | None

model_config = {'extra': 'forbid'}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

classmethod validate_quantification_id_input(v: Any) → Any

classmethod validate_non_empty_identifier(v: int) → int

classmethod validate_analytical_error(v: float | None) → float | None