gwin.models.base_data module

Base classes for models with data.

class gwin.models.base_data.BaseDataModel(variable_params, data, waveform_generator, waveform_transforms=None, **kwargs)

Bases: gwin.models.base.BaseModel

Base class for models that require data and a waveform generator.

This adds propeties for the log of the likelihood that the data contain noise, lognl, and the log likelihood ratio loglr.

Classes that inherit from this class must define _loglr and _lognl functions, in addition to the _loglikelihood requirement inherited from BaseModel.

Parameters:

• variable_params ((tuple of) string(s)) – A tuple of parameter names that will be varied.
• data (dict) – A dictionary of data, in which the keys are the detector names and the values are the data.
• waveform_generator (generator class) – A generator class that creates waveforms.
• waveform_transforms (list, optional) – List of transforms to use to go from the variable args to parameters understood by the waveform generator.
• **kwargs – All other keyword arguments are passed to BaseModel.

waveform_generator

dict – The waveform generator that the class was initialized with.

data

dict – The data that the class was initialized with.

Properties

----------

lognl

Returns the log likelihood of the noise.

loglr

Returns the log of the likelihood ratio.

logplr

Returns the log of the prior-weighted likelihood ratio.

See ``BaseModel`` for additional attributes and properties.

data

Returns the data that was set.

classmethod from_config(cp, data, delta_f=None, delta_t=None, gates=None, recalibration=None, **kwargs)

Initializes an instance of this class from the given config file.

Parameters:

• cp (WorkflowConfigParser) – Config file parser to read.
• data (dict) – A dictionary of data, in which the keys are the detector names and the values are the data. This is not retrieved from the config file, and so must be provided.
• delta_f (float) – The frequency spacing of the data; needed for waveform generation.
• delta_t (float) – The time spacing of the data; needed for time-domain waveform generators.
• recalibration (dict of pycbc.calibration.Recalibrate, optional) – Dictionary of detectors -> recalibration class instances for recalibrating data.
• gates (dict of tuples, optional) – Dictionary of detectors -> tuples of specifying gate times. The sort of thing returned by pycbc.gate.gates_from_cli.
• **kwargs – All additional keyword arguments are passed to the class. Any provided keyword will over ride what is in the config file.

loglr

The log likelihood ratio at the current parameters.

This will initially try to return the current_stats.loglr. If that raises an AttributeError, will call _loglr` to calculate it and store it to current_stats.

lognl

The log likelihood of the model assuming the data is noise.

This will initially try to return the current_stats.lognl. If that raises an AttributeError, will call _lognl` to calculate it and store it to current_stats.

logplr

Returns the log of the prior-weighted likelihood ratio at the current parameter values.

The logprior is calculated first. If the logprior returns -inf (possibly indicating a non-physical point), then loglr is not called.

waveform_generator

Returns the waveform generator that was set.