Model

class pyhf.pdf.Model(spec, modifier_set=None, batch_size=None, validate: bool = True, clip_sample_data: float | None = None, clip_bin_data: float | None = None, **config_kwargs)[source]

Bases: object

The main pyhf model class.

__init__(spec, modifier_set=None, batch_size=None, validate: bool = True, clip_sample_data: float | None = None, clip_bin_data: float | None = None, **config_kwargs)[source]

Construct a HistFactory Model.

Parameters:

spec (jsonable) – The HistFactory JSON specification
batch_size (None or int) – Number of simultaneous (batched) Models to compute.
validate (bool) – Whether to validate against a JSON schema
clip_sample_data (None or float) – Clip the minimum value of expected data by-sample. Default is no clipping.
clip_bin_data (None or float) – Clip the minimum value of expected data by-bin. Default is no clipping.
config_kwargs – Possible keyword arguments for the model configuration

Returns:

The Model instance.

Return type:

model (Model)

Attributes

config: The _ModelConfig instance for the model.

nominal_rates: Nominal value of bin rates of the main model.

Methods

constraint_logpdf(auxdata, pars)[source]

Compute the log value of the constraint pdf.

Parameters:

auxdata (tensor) – The auxiliary measurement data
pars (tensor) – The parameter values

Returns:

The log density value

Return type:

Tensor

expected_actualdata(pars)[source]

Compute the expected value of the main model.

Parameters:: pars (tensor) – The parameter values
Returns:: The expected data of the main model (no auxiliary data)
Return type:: Tensor

expected_auxdata(pars)[source]

Compute the expected value of the auxiliary measurements.

Parameters:: pars (tensor) – The parameter values
Returns:: The expected data of the auxiliary pdf
Return type:: Tensor

expected_data(pars, include_auxdata=True)[source]

Compute the expected value of the main model

Parameters:: pars (tensor) – The parameter values
Returns:: The expected data of the main and auxiliary model
Return type:: Tensor

logpdf(pars, data)[source]

Compute the log value of the full density.

Parameters:

pars (tensor) – The parameter values
data (tensor) – The measurement data

Returns:

The log density value

Return type:

Tensor

mainlogpdf(maindata, pars)[source]

Compute the log value of the main term.

Parameters:

maindata (tensor) – The main measurement data
pars (tensor) – The parameter values

Returns:

The log density value

Return type:

Tensor

make_pdf(pars)[source]

Construct a pdf object for a given set of parameter values.

Parameters:: pars (tensor) – The model parameters
Returns:: A distribution object implementing the main measurement pdf of HistFactory
Return type:: pdf

modifications(pars)[source]: The modifications applied to the _MainModel. See pyhf.pdf._MainModel.modifications() for details.

pdf(pars, data)[source]

Compute the density at a given observed point in data space of the full model.

Parameters:

pars (tensor) – The parameter values
data (tensor) – The measurement data

Returns:

The density value

Return type:

Tensor