pyhs3.distributions.MixtureDist¶

class pyhs3.distributions.MixtureDist(**data)[source]¶

Mixture of probability distributions.

Implements a weighted combination of multiple distributions following ROOT’s RooAddPdf. Supports both N and N-1 coefficient configurations where $N$ represents number of distributions (summands):

N-1 coefficients:

\[f(x) = \sum_{i=1}^{n-1} c_i \cdot f_i(x) + (1 - \sum_{i=1}^{n-1} c_i) \cdot f_n(x)\]

N coefficients:

\[f(x) = \frac{\sum_{i=1}^{n} c_i \cdot f_i(x)}{\sum_{i=1}^{n} c_i}\]

N coefficients with `ref_coef_norm`:

\[f(x) = \frac{\sum_{i=1}^{n} c_i \cdot f_i(x)}{\sum_{j \in \text{ref\_coef\_norm}} c_j}\]

log_prob_terms() is the log-space-safe path for an extended mixture: it builds log(Σcᵢfᵢ) via logsumexp over each summand’s own log-space expression rather than pt.log of the probability-space sum, so it stays finite even when every component underflows to 0.0. The base-class log_expression() (used for non-extended mixtures and standalone evaluation) still round-trips through probability-space component tensors and cannot be converted until composite distributions receive log-space component context (see https://github.com/scipp-atlas/pyhs3/issues/254, phase 4).

Parameters:

coefficients (list[str]) – Names of coefficient parameters.
summands (list[str]) – Names of component distributions.
extended (bool) – Whether the mixture is extended (affects normalization). Must be True for N coefficients, False for N-1 coefficients.
ref_coef_norm (list[str] | None) – Optional list of coefficient names for custom normalization. Only valid when using N coefficients (extended=True).

ROOT Reference

RooAddPdf

Parameters:: data (Any)

__init__(**data)¶

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:: data (Any)

Methods

`__init__`(**data)	Create a new model by parsing and validating input data from keyword arguments.
`construct`([_fields_set])
`copy`(*[, include, exclude, update, deep])	Returns a copy of the model.
`dict`(*[, include, exclude, by_alias, ...])
`expected_yield`(context)	Compute the total expected yield nu in the extended case.
`expression`(context)	Evaluate and return a named PyTensor expression.
`extended_likelihood`(_context[, _data])	Extended likelihood contribution in normal space.
`from_orm`(obj)
`get_parameter_list`(context, param_key)	Reconstruct a parameter list from flattened indexed keys.
`json`(*[, include, exclude, by_alias, ...])
`likelihood`(context)	Builds a symbolic expression for the mixture distribution.
`log_expression`(context)	Log-probability combining main likelihood with extended terms.
`log_likelihood`(context)	Log-space counterpart of `likelihood()`.
`log_prob_terms`(expressions, log_expressions, ...)	Extended-likelihood contributions for the mixture.
`model_construct`([_fields_set])	Creates a new instance of the Model class with validated data.
`model_copy`(*[, update, deep])	!!! abstract "Usage Documentation"
`model_dump`(*[, mode, include, exclude, ...])	!!! abstract "Usage Documentation"
`model_dump_json`(*[, indent, ensure_ascii, ...])	!!! abstract "Usage Documentation"
`model_json_schema`([by_alias, ref_template, ...])	Generates a JSON schema for a model class.
`model_parametrized_name`(params)	Compute the class name for parametrizations of generic classes.
`model_post_init`(context, /)	This function is meant to behave like a BaseModel method to initialise private attributes.
`model_rebuild`(*[, force, raise_errors, ...])	Try to rebuild the pydantic-core schema for the model.
`model_validate`(obj, *[, strict, extra, ...])	Validate a pydantic model instance.
`model_validate_json`(json_data, *[, strict, ...])	!!! abstract "Usage Documentation"
`model_validate_strings`(obj, *[, strict, ...])	Validate the given object with string data against the Pydantic model.
`normalization_expression`(_context, ...)	Return the antiderivative expression, or None for numerical fallback.
`parse_file`(path, *[, content_type, ...])
`parse_obj`(obj)
`parse_raw`(b, *[, content_type, encoding, ...])
`process_parameter`(param_key)	Process a single parameter that can be either a string reference or numeric value.
`process_parameter_list`(param_key)	Process a list parameter containing mixed string references and numeric values.
`schema`([by_alias, ref_template])
`schema_json`(*[, by_alias, ref_template])
`serialize_model`(handler)	Do not serialize ref_coef_norm if it is unspecified (None).
`serialize_ref_coef_norm`(ref_coef_norm)	Convert list back to comma-separated string for serialization.
`split_comma_separated_ref_coef_norm`(v)	Convert comma-separated string to list for ref_coef_norm.
`unnormalized_expression`(context)	Return the unnormalized mixture Σcᵢfᵢ (before dividing by Σcᵢ).
`update_forward_refs`(**localns)
`validate`(value)
`validate_coefficient_count`(coefficients, info)	Validate that coefficient count matches summand count appropriately.
`validate_extended_matches_coefficients`(...)	Validate that extended matches coefficient configuration.
`validate_ref_coef_norm_usage`(ref_coef_norm, info)	Validate that ref_coef_norm is only used with N=N coefficient case.

Attributes

`constants`	Dictionary of PyTensor constants generated from numeric field values.
`model_computed_fields`
`model_config`	Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].
`model_extra`	Get extra fields set during validation.
`model_fields`
`model_fields_set`	Returns the set of fields that have been explicitly set on this model instance.
`parameters`	Set of parameter names this component depends on.
`type`
`summands`
`coefficients`
`extended`
`ref_coef_norm`
`name`

classmethod MixtureDist.construct(_fields_set=None, **values)¶

Parameters:

_fields_set (set[str] | None)
values (Any)

Return type:

Self

MixtureDist.copy(*, include=None, exclude=None, update=None, deep=False)¶

Returns a copy of the model.

!!! warning “Deprecated”: This method is now deprecated; use model_copy instead.

If you need include or exclude, use:

`python {test="skip" lint="skip"} data = self.model_dump(include=include, exclude=exclude, round_trip=True) data = {**data, **(update or {})} copied = self.model_validate(data) `

Parameters:

include (Set[int] | Set[str] | Mapping[int, Any] | Mapping[str, Any] | None) – Optional set or mapping specifying which fields to include in the copied model.
exclude (Set[int] | Set[str] | Mapping[int, Any] | Mapping[str, Any] | None) – Optional set or mapping specifying which fields to exclude in the copied model.
update (Dict[str, Any] | None) – Optional dictionary of field-value pairs to override field values in the copied model.
deep (bool) – If True, the values of fields that are Pydantic models will be deep-copied.

Return type:

Self

Returns:

A copy of the model with included, excluded and updated fields as specified.

MixtureDist.dict(*, include=None, exclude=None, by_alias=False, exclude_unset=False, exclude_defaults=False, exclude_none=False)¶

Parameters:

include (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None)
exclude (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None)
by_alias (bool)
exclude_unset (bool)
exclude_defaults (bool)
exclude_none (bool)

Return type:

Dict[str, Any]

MixtureDist.expected_yield(context)[source]¶

Compute the total expected yield nu in the extended case.

N coefficients case: nu = sum(coefficients) or sum(ref_coef_norm) if specified
N-1 coefficients case: not defined (extended=False always)

Parameters:: context (Context) – Mapping of names to pytensor variables
Return type:: TypeAliasType
Returns:: Expected yield (nu) for extended likelihood
Raises:: RuntimeError – If called on non-extended PDF

MixtureDist.expression(context)¶

Evaluate and return a named PyTensor expression.

This is a template method that calls _expression() to get the result, then names the result before returning.

_expression() may return a tensor it does not own — for example a single-summand SumFunction or a single-factor ProductFunction returns context[param] directly, which is the same tensor object stored in model.parameters. Renaming such a shared tensor in place would silently corrupt the parameter it borrows (e.g. model.parameters['mu'] would be renamed to the function’s name). To avoid this, a result that already carries a different name is wrapped in a fresh identity copy before being renamed; freshly built (unnamed) expressions are named in place as before so that compilation and graph_summary keep working.

Parameters:: context (Context) – Mapping of names to PyTensor variables
Return type:: TypeAliasType
Returns:: Named PyTensor expression representing the component

MixtureDist.extended_likelihood(_context, _data=None)¶

Extended likelihood contribution in normal space.

Returns additional likelihood terms for extended ML fitting. Override only when the extended terms belong in the distribution’s per-event density, like constraint terms (HistFactory). Terms that enter the likelihood once per channel (e.g. the Poisson yield term of an extended MixtureDist, which involves the observed event count) must not use this hook: _expression() multiplies the result into the density, so Model.log_prob would count it once per event when summing over data. Such terms are assembled once per channel by Model.log_prob, which owns the channel-dataset pairing.

Default: no contribution (returns 1.0 in normal space).

Parameters:

context – Mapping of names to pytensor variables
data – Optional data tensor for data-dependent terms
_context (Context)
_data (TypeAliasType | None)

Returns:

Likelihood contribution (default: 1.0 = no contribution)

Return type:

TensorVar

classmethod MixtureDist.from_orm(obj)¶

Parameters:: obj (Any)
Return type:: Self

MixtureDist.get_parameter_list(context, param_key)¶

Reconstruct a parameter list from flattened indexed keys.

Used to recover the original list structure from the indexed parameter mapping created by process_parameter_list().

Parameters:

context (Context) – The context containing parameter values mapped by name.
param_key (str) – The base parameter key (e.g., “factors”).

Return type:

list[TypeAliasType]

Returns:

List of parameter values in original order.

Example

>>> from typing import Literal
>>> class TestEvaluable(Evaluable):
...     type: Literal["test"] = "test"
...     factors: list[str | float]
...     def _expression(self, _: Context) -> TensorVar:
...         return None
>>>
>>> eval1 = TestEvaluable(name="test", factors=["a", 1.0, "b"])
>>> context = {
...     "a": "tensor_a",
...     "constant_test_factors[1]": "tensor_1",
...     "b": "tensor_b"
... }
>>> eval1.get_parameter_list(context, "factors")
['tensor_a', 'tensor_1', 'tensor_b']

MixtureDist.json(*, include=None, exclude=None, by_alias=False, exclude_unset=False, exclude_defaults=False, exclude_none=False, encoder=PydanticUndefined, models_as_dict=PydanticUndefined, **dumps_kwargs)¶

Parameters:

include (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None)
exclude (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None)
by_alias (bool)
exclude_unset (bool)
exclude_defaults (bool)
exclude_none (bool)
encoder (Callable[[Any], Any] | None)
models_as_dict (bool)
dumps_kwargs (Any)

Return type:

str

MixtureDist.likelihood(context)[source]¶

Builds a symbolic expression for the mixture distribution.

Handles both N and N-1 coefficient cases: - N-1 coefficients: Traditional approach with automatic normalization - N coefficients: Direct summation with optional custom normalization

Parameters:: context (dict) – Mapping of names to pytensor variables.
Returns:: Symbolic representation of the mixture PDF.
Return type:: pytensor.tensor.variable.TensorVariable

MixtureDist.log_expression(context)¶

Log-probability combining main likelihood with extended terms.

Returns log_likelihood() plus log(extended_likelihood()), minus the log of the normalization integral where normalization applies. This is mathematically equivalent to log(likelihood * extended_likelihood) but starts from log_likelihood() so that a distribution’s analytic log form (e.g. Gaussian, Poisson, LogNormal) is used directly instead of round- tripping through a probability value that can underflow to 0.0.

Normalization itself stays probability-space: the normalization integral (whether analytic or Gauss-Legendre quadrature) is a sum of positive terms and is far less underflow-prone than the raw density, so log(integral) is subtracted rather than computed as its own analytic log form.

All distributions are automatically normalized over observables present in the context, unless explicitly opted out via _normalizable = False.

PyTensor handles optimization and simplification automatically.

Parameters:: context (Context) – Mapping of names to pytensor variables
Returns:: Log-probability density
Return type:: TensorVar

MixtureDist.log_likelihood(context)¶

Log-space counterpart of likelihood().

Default implementation takes the log of the probability-space likelihood. Override with an analytic log-space form wherever one exists (e.g. Gaussian, Poisson, LogNormal): the default round-trips through a probability value that can underflow to 0.0 for parameter points far from the mode, at which point pt.log(0.0) is -inf even though the true log-density is finite.

Parameters:: context (Context) – Mapping of names to pytensor variables
Returns:: Log of the main probability density
Return type:: TensorVar

MixtureDist.log_prob_terms(expressions, log_expressions, distributions)[source]¶

Extended-likelihood contributions for the mixture.

For an extended mixture the per-channel log-likelihood is built from the unnormalized mixture and the expected yield:

Σⱼ log(Σᵢcᵢfᵢ(xⱼ)) - nu

which simplifies from Σⱼ log(Σᵢcᵢfᵢ(xⱼ)/nu) + N·log(nu) - nu (the log(nu) terms cancel). This is algebraically identical to the full extended likelihood but avoids introducing pt.log(nu) into the graph, preventing costly optimizer rewrite cascades triggered by log(a/b) + N·log(b). For weighted data this form also reproduces RooFit’s sum-of-weights convention: weighting the log(Σᵢcᵢfᵢ) term gives Σⱼwⱼ·log(PDF) + (Σwⱼ)·log(nu) - nu.

log(Σᵢcᵢfᵢ(xⱼ)) is evaluated as logsumexp(log cᵢ + log fᵢ(xⱼ)) using each summand’s own log-space expression (log_expressions) rather than pt.log of the probability-space sum (_cached_unnorm_expr). fᵢ(xⱼ) can underflow to 0.0 in float64 for xⱼ far in a component’s tail, which would make every cᵢfᵢ(xⱼ) — and hence their sum — exactly 0.0 and pt.log of it -inf, even though log fᵢ(xⱼ) itself is finite. Building the sum in log space keeps it finite in that regime. The log cᵢ + log fᵢ(xⱼ) terms are stacked on a new leading axis (shape (n_summands, N, M) for N events and parameter-batch size M, matching the per-event shape documented on LogProbTerms) and reduced with _stable_logsumexp() over that axis, giving the (N, M) per-event term. _stable_logsumexp() is used instead of pt.logsumexp because the latter’s stability is an emergent property of a graph rewrite that only runs under FAST_RUN (see _stable_logsumexp()’s docstring and https://github.com/scipp-atlas/pyhs3/issues/277). This assumes cᵢ ≥ 0, matching the probabilistic meaning of a mixture weight; MixtureDist does not validate coefficient sign, so a negative cᵢ produces NaN here (pt.log of a negative number) just as pt.log(Σᵢcᵢfᵢ) would if the probability-space sum itself were negative.

The data-only -log(N!) constant is omitted and N_eff = Σwⱼ is used for weighted data; both are RooFit conventions adopted because HS3 does not specify the extended term exactly. See https://github.com/hep-statistics-serialization-standard/hep-statistics-serialization-standard/issues/91 and https://github.com/scipp-atlas/pyhs3/issues/241.

Non-extended mixtures contribute the default per-event log(PDF).

Parameters:

expressions (Mapping[str, TypeAliasType])
log_expressions (Mapping[str, TypeAliasType])
distributions (Distributions)

Return type:

LogProbTerms

classmethod MixtureDist.model_construct(_fields_set=None, **values)¶

Creates a new instance of the Model class with validated data.

Creates a new model setting __dict__ and __pydantic_fields_set__ from trusted or pre-validated data. Default values are respected, but no other validation is performed.

!!! note: model_construct() generally respects the model_config.extra setting on the provided model. That is, if model_config.extra == ‘allow’, then all extra passed values are added to the model instance’s __dict__ and __pydantic_extra__ fields. If model_config.extra == ‘ignore’ (the default), then all extra passed values are ignored. Because no validation is performed with a call to model_construct(), having model_config.extra == ‘forbid’ does not result in an error if extra values are passed, but they will be ignored.

Parameters:

_fields_set (set[str] | None) – A set of field names that were originally explicitly set during instantiation. If provided, this is directly used for the [model_fields_set][pydantic.BaseModel.model_fields_set] attribute. Otherwise, the field names from the values argument will be used.
values (Any) – Trusted or pre-validated data dictionary.

Return type:

Self

Returns:

A new instance of the Model class with validated data.

MixtureDist.model_copy(*, update=None, deep=False)¶

!!! abstract “Usage Documentation”: [model_copy](../concepts/models.md#model-copy)

Returns a copy of the model.

!!! note: The underlying instance’s [__dict__][object.__dict__] attribute is copied. This might have unexpected side effects if you store anything in it, on top of the model fields (e.g. the value of [cached properties][functools.cached_property]).

Parameters:

update (Mapping[str, Any] | None) – Values to change/add in the new model. Note: the data is not validated before creating the new model. You should trust this data.
deep (bool) – Set to True to make a deep copy of the model.

Return type:

Self

Returns:

New model instance.

MixtureDist.model_dump(*, mode='python', include=None, exclude=None, context=None, by_alias=None, exclude_unset=False, exclude_defaults=False, exclude_none=False, exclude_computed_fields=False, round_trip=False, warnings=True, fallback=None, serialize_as_any=False)¶

!!! abstract “Usage Documentation”: [model_dump](../concepts/serialization.md#python-mode)

Generate a dictionary representation of the model, optionally specifying which fields to include or exclude.

Parameters:

mode (Literal['json', 'python'] | str) – The mode in which to_python should run. If mode is ‘json’, the output will only contain JSON serializable types. If mode is ‘python’, the output may contain non-JSON-serializable Python objects.
include (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None) – A set of fields to include in the output.
exclude (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None) – A set of fields to exclude from the output.
context (Any | None) – Additional context to pass to the serializer.
by_alias (bool | None) – Whether to use the field’s alias in the dictionary key if defined.
exclude_unset (bool) – Whether to exclude fields that have not been explicitly set.
exclude_defaults (bool) – Whether to exclude fields that are set to their default value.
exclude_none (bool) – Whether to exclude fields that have a value of None.
exclude_computed_fields (bool) – Whether to exclude computed fields. While this can be useful for round-tripping, it is usually recommended to use the dedicated round_trip parameter instead.
round_trip (bool) – If True, dumped values should be valid as input for non-idempotent types such as Json[T].
warnings (bool | Literal['none', 'warn', 'error']) – How to handle serialization errors. False/”none” ignores them, True/”warn” logs errors, “error” raises a [PydanticSerializationError][pydantic_core.PydanticSerializationError].
fallback (Callable[[Any], Any] | None) – A function to call when an unknown value is encountered. If not provided, a [PydanticSerializationError][pydantic_core.PydanticSerializationError] error is raised.
serialize_as_any (bool) – Whether to serialize fields with duck-typing serialization behavior.

Return type:

dict[str, Any]

Returns:

A dictionary representation of the model.

MixtureDist.model_dump_json(*, indent=None, ensure_ascii=False, include=None, exclude=None, context=None, by_alias=None, exclude_unset=False, exclude_defaults=False, exclude_none=False, exclude_computed_fields=False, round_trip=False, warnings=True, fallback=None, serialize_as_any=False)¶

!!! abstract “Usage Documentation”: [model_dump_json](../concepts/serialization.md#json-mode)

Generates a JSON representation of the model using Pydantic’s to_json method.

Parameters:

indent (int | None) – Indentation to use in the JSON output. If None is passed, the output will be compact.
ensure_ascii (bool) – If True, the output is guaranteed to have all incoming non-ASCII characters escaped. If False (the default), these characters will be output as-is.
include (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None) – Field(s) to include in the JSON output.
exclude (set[int] | set[str] | Mapping[int, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | Mapping[str, set[int] | set[str] | Mapping[int, IncEx | bool] | Mapping[str, IncEx | bool] | bool] | None) – Field(s) to exclude from the JSON output.
context (Any | None) – Additional context to pass to the serializer.
by_alias (bool | None) – Whether to serialize using field aliases.
exclude_unset (bool) – Whether to exclude fields that have not been explicitly set.
exclude_defaults (bool) – Whether to exclude fields that are set to their default value.
exclude_none (bool) – Whether to exclude fields that have a value of None.
exclude_computed_fields (bool) – Whether to exclude computed fields. While this can be useful for round-tripping, it is usually recommended to use the dedicated round_trip parameter instead.
round_trip (bool) – If True, dumped values should be valid as input for non-idempotent types such as Json[T].
warnings (bool | Literal['none', 'warn', 'error']) – How to handle serialization errors. False/”none” ignores them, True/”warn” logs errors, “error” raises a [PydanticSerializationError][pydantic_core.PydanticSerializationError].
fallback (Callable[[Any], Any] | None) – A function to call when an unknown value is encountered. If not provided, a [PydanticSerializationError][pydantic_core.PydanticSerializationError] error is raised.
serialize_as_any (bool) – Whether to serialize fields with duck-typing serialization behavior.

Return type:

str

Returns:

A JSON string representation of the model.

classmethod MixtureDist.model_json_schema(by_alias=True, ref_template='#/$defs/{model}', schema_generator=<class 'pydantic.json_schema.GenerateJsonSchema'>, mode='validation', *, union_format='any_of')¶

Generates a JSON schema for a model class.

Parameters:

by_alias (bool) – Whether to use attribute aliases or not.
ref_template (str) – The reference template.
union_format (Literal['any_of', 'primitive_type_array']) –
The format to use when combining schemas from unions together. Can be one of:
- ’any_of’: Use the [anyOf](https://json-schema.org/understanding-json-schema/reference/combining#anyOf)
keyword to combine schemas (the default). - ‘primitive_type_array’: Use the [type](https://json-schema.org/understanding-json-schema/reference/type) keyword as an array of strings, containing each type of the combination. If any of the schemas is not a primitive type (string, boolean, null, integer or number) or contains constraints/metadata, falls back to any_of.
schema_generator (type[GenerateJsonSchema]) – To override the logic used to generate the JSON schema, as a subclass of GenerateJsonSchema with your desired modifications
mode (Literal['validation', 'serialization']) – The mode in which to generate the schema.

Return type:

dict[str, Any]

Returns:

The JSON schema for the given model class.

classmethod MixtureDist.model_parametrized_name(params)¶

Compute the class name for parametrizations of generic classes.

This method can be overridden to achieve a custom naming scheme for generic BaseModels.

Parameters:: params (tuple[type[Any], ...]) – Tuple of types of the class. Given a generic class Model with 2 type variables and a concrete model Model[str, int], the value (str, int) would be passed to params.
Return type:: str
Returns:: String representing the new class where params are passed to cls as type variables.
Raises:: TypeError – Raised when trying to generate concrete names for non-generic models.

MixtureDist.model_post_init(context, /)¶

This function is meant to behave like a BaseModel method to initialise private attributes.

It takes context as an argument since that’s what pydantic-core passes when calling it.

Parameters:

self (BaseModel) – The BaseModel instance.
context (Any) – The context.

Return type:

None

classmethod MixtureDist.model_rebuild(*, force=False, raise_errors=True, _parent_namespace_depth=2, _types_namespace=None)¶

Try to rebuild the pydantic-core schema for the model.

This may be necessary when one of the annotations is a ForwardRef which could not be resolved during the initial attempt to build the schema, and automatic rebuilding fails.

Parameters:

force (bool) – Whether to force the rebuilding of the model schema, defaults to False.
raise_errors (bool) – Whether to raise errors, defaults to True.
_parent_namespace_depth (int) – The depth level of the parent namespace, defaults to 2.
_types_namespace (Mapping[str, Any] | None) – The types namespace, defaults to None.

Return type:

bool | None

Returns:

Returns None if the schema is already “complete” and rebuilding was not required. If rebuilding _was_ required, returns True if rebuilding was successful, otherwise False.

classmethod MixtureDist.model_validate(obj, *, strict=None, extra=None, from_attributes=None, context=None, by_alias=None, by_name=None)¶

Validate a pydantic model instance.

Parameters:

obj (Any) – The object to validate.
strict (bool | None) – Whether to enforce types strictly.
extra (Literal['allow', 'ignore', 'forbid'] | None) – Whether to ignore, allow, or forbid extra data during model validation. See the [extra configuration value][pydantic.ConfigDict.extra] for details.
from_attributes (bool | None) – Whether to extract data from object attributes.
context (Any | None) – Additional context to pass to the validator.
by_alias (bool | None) – Whether to use the field’s alias when validating against the provided input data.
by_name (bool | None) – Whether to use the field’s name when validating against the provided input data.

Raises:

ValidationError – If the object could not be validated.

Return type:

Self

Returns:

The validated model instance.

classmethod MixtureDist.model_validate_json(json_data, *, strict=None, extra=None, context=None, by_alias=None, by_name=None)¶

!!! abstract “Usage Documentation”: [JSON Parsing](../concepts/json.md#json-parsing)

Validate the given JSON data against the Pydantic model.

Parameters:

json_data (str | bytes | bytearray) – The JSON data to validate.
strict (bool | None) – Whether to enforce types strictly.
extra (Literal['allow', 'ignore', 'forbid'] | None) – Whether to ignore, allow, or forbid extra data during model validation. See the [extra configuration value][pydantic.ConfigDict.extra] for details.
context (Any | None) – Extra variables to pass to the validator.
by_alias (bool | None) – Whether to use the field’s alias when validating against the provided input data.
by_name (bool | None) – Whether to use the field’s name when validating against the provided input data.

Return type:

Self

Returns:

The validated Pydantic model.

Raises:

ValidationError – If json_data is not a JSON string or the object could not be validated.

classmethod MixtureDist.model_validate_strings(obj, *, strict=None, extra=None, context=None, by_alias=None, by_name=None)¶

Validate the given object with string data against the Pydantic model.

Parameters:

obj (Any) – The object containing string data to validate.
strict (bool | None) – Whether to enforce types strictly.
extra (Literal['allow', 'ignore', 'forbid'] | None) – Whether to ignore, allow, or forbid extra data during model validation. See the [extra configuration value][pydantic.ConfigDict.extra] for details.
context (Any | None) – Extra variables to pass to the validator.
by_alias (bool | None) – Whether to use the field’s alias when validating against the provided input data.
by_name (bool | None) – Whether to use the field’s name when validating against the provided input data.

Return type:

Self

Returns:

The validated Pydantic model.

MixtureDist.normalization_expression(_context, _observable_name)¶

Return the antiderivative expression, or None for numerical fallback.

Override in subclasses to provide analytical normalization. The returned expression should be the antiderivative F(x) such that the integral ∫f(x)dx from a to b equals F(b) - F(a).

Parameters:

context – Mapping of names to pytensor variables
observable_name – Name of the observable to integrate over
_context (Context)
_observable_name (str)

Return type:

TypeAliasType | None

Returns:

Symbolic antiderivative expression, or None for numerical fallback.

classmethod MixtureDist.parse_file(path, *, content_type=None, encoding='utf8', proto=None, allow_pickle=False)¶

Parameters:

path (str | Path)
content_type (str | None)
encoding (str)
proto (Protocol | None)
allow_pickle (bool)

Return type:

Self

classmethod MixtureDist.parse_obj(obj)¶

Parameters:: obj (Any)
Return type:: Self

classmethod MixtureDist.parse_raw(b, *, content_type=None, encoding='utf8', proto=None, allow_pickle=False)¶

Parameters:

b (str | bytes)
content_type (str | None)
encoding (str)
proto (Protocol | None)
allow_pickle (bool)

Return type:

Self

MixtureDist.process_parameter(param_key)¶

Process a single parameter that can be either a string reference or numeric value.

For numeric values, generates a unique constant name. For string values, returns the value as-is.

Parameters:

param_key (str) – The parameter field name to process (e.g., “mean”, “sigma”).

Returns:

processed_name: Either the original string value or a generated constant name
numeric_value: The numeric value if input was numeric, None otherwise

Return type:

Tuple containing

Example

>>> from typing import Literal
>>> class TestEvaluable(Evaluable):
...     type: Literal["test"] = "test"
...     some_param: str | float
...     def _expression(self, _: Context) -> TensorVar:
...         return None
>>>
>>> # String parameter
>>> eval1 = TestEvaluable(name="test1", some_param="alpha")
>>> eval1.process_parameter("some_param")
('alpha', None)

>>> # Numeric parameter
>>> eval2 = TestEvaluable(name="test2", some_param=1.5)
>>> eval2.process_parameter("some_param")
('constant_test2_some_param', 1.5)

MixtureDist.process_parameter_list(param_key)¶

Process a list parameter containing mixed string references and numeric values.

For numeric values, generates indexed unique names and stores the values. For string values, returns the values as-is. Also updates internal parameter mapping with indexed keys.

Parameters:

param_key (str) – The parameter field name to process (e.g., “factors”, “coefficients”).

Returns:

processed_names: List of parameter names (original strings or generated constant names)
numeric_values: List of numeric values (None for string entries)

Return type:

Tuple containing

Example

>>> from typing import Literal
>>> class TestEvaluable(Evaluable):
...     type: Literal["test"] = "test"
...     factors: list[str | float]
...     def _expression(self, _: Context) -> TensorVar:
...         return None
>>>
>>> eval1 = TestEvaluable(name="test", factors=["param1", 2.0, "param2"])
>>> names, values = eval1.process_parameter_list("factors")
>>> names
['param1', 'constant_test_factors[1]', 'param2']
>>> values
[None, 2.0, None]

classmethod MixtureDist.schema(by_alias=True, ref_template='#/$defs/{model}')¶

Parameters:

by_alias (bool)
ref_template (str)

Return type:

Dict[str, Any]

classmethod MixtureDist.schema_json(*, by_alias=True, ref_template='#/$defs/{model}', **dumps_kwargs)¶

Parameters:

by_alias (bool)
ref_template (str)
dumps_kwargs (Any)

Return type:

str

MixtureDist.serialize_model(handler)[source]¶

Do not serialize ref_coef_norm if it is unspecified (None).

Parameters:: handler (Callable[[Any], Any])
Return type:: Any

MixtureDist.serialize_ref_coef_norm(ref_coef_norm)[source]¶

Convert list back to comma-separated string for serialization.

Parameters:: ref_coef_norm (list[str] | None)
Return type:: str | None

classmethod MixtureDist.split_comma_separated_ref_coef_norm(v)[source]¶

Convert comma-separated string to list for ref_coef_norm.

Parameters:: v (object)
Return type:: object

MixtureDist.unnormalized_expression(context)[source]¶

Return the unnormalized mixture Σcᵢfᵢ (before dividing by Σcᵢ).

After likelihood() has been called this just returns the cached intermediate result. This is a probability-space quantity: for the per-event log-likelihood, log_prob_terms() builds log(Σcᵢfᵢ(xⱼ)) directly from each summand’s own log-space expression via logsumexp instead of taking pt.log of this sum, so that the result stays finite where the probability-space fᵢ(xⱼ) underflow to 0.0. This method remains available for callers wanting the raw probability-space value itself.

The Poisson yield term enters the likelihood once per channel and involves the observed event count, so it is assembled by pyhs3.Model.log_prob() (which owns the channel-dataset pairing) rather than via extended_likelihood() (whose result is multiplied into the per-event density and would be overcounted when summing over events).

Parameters:: context (Context)
Return type:: TypeAliasType

classmethod MixtureDist.update_forward_refs(**localns)¶

Parameters:: localns (Any)
Return type:: None

classmethod MixtureDist.validate(value)¶

Parameters:: value (Any)
Return type:: Self

classmethod MixtureDist.validate_coefficient_count(coefficients, info)[source]¶

Validate that coefficient count matches summand count appropriately.

Parameters:

coefficients (list[str])
info (ValidationInfo)

Return type:

list[str]

classmethod MixtureDist.validate_extended_matches_coefficients(extended, info)[source]¶

Validate that extended matches coefficient configuration.

Parameters:

extended (bool)
info (ValidationInfo)

Return type:

bool

classmethod MixtureDist.validate_ref_coef_norm_usage(ref_coef_norm, info)[source]¶

Validate that ref_coef_norm is only used with N=N coefficient case.

Parameters:

ref_coef_norm (list[str] | None)
info (ValidationInfo)

Return type:

list[str] | None

MixtureDist.constants¶

Dictionary of PyTensor constants generated from numeric field values.

Returns:: Mapping from generated constant names to PyTensor constant tensors. Empty if all fields are string references.

MixtureDist.model_computed_fields = {}¶

MixtureDist.model_config: ClassVar[ConfigDict] = {'serialize_by_alias': True}¶: Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

MixtureDist.model_extra¶

Get extra fields set during validation.

Returns:: A dictionary of extra fields, or None if config.extra is not set to “allow”.

MixtureDist.model_fields = {'coefficients': FieldInfo(annotation=list[str], required=True), 'extended': FieldInfo(annotation=bool, required=False, default=False), 'name': FieldInfo(annotation=str, required=True, json_schema_extra={'preprocess': False}), 'ref_coef_norm': FieldInfo(annotation=Union[list[str], NoneType], required=False, default=None, json_schema_extra={'preprocess': False}), 'summands': FieldInfo(annotation=list[str], required=True), 'type': FieldInfo(annotation=Literal['mixture_dist'], required=False, default='mixture_dist')}¶

MixtureDist.model_fields_set¶

Returns the set of fields that have been explicitly set on this model instance.

Returns:

A set of strings representing the fields that have been set,: i.e. that were not filled from defaults.

MixtureDist.parameters¶

Set of parameter names this component depends on.

Returns:: Set of parameter names, including both string references and generated constant names for numeric values.

MixtureDist.type: Literal['mixture_dist']¶

MixtureDist.summands: list[str]¶

MixtureDist.coefficients: list[str]¶

MixtureDist.extended: bool¶

MixtureDist.ref_coef_norm: list[str] | None¶

MixtureDist.name: str¶