GWForge.population.mass
=======================

.. py:module:: GWForge.population.mass


Attributes
----------

.. autoapisummary::

   GWForge.population.mass.choices
   GWForge.population.mass.sampler_choices
   GWForge.population.mass.BGP_PARAMETERS
   GWForge.population.mass.POWER_LAW_PEAK_PARAMETERS
   GWForge.population.mass.DEFAULT_PARAMETERS
   GWForge.population.mass.CONDITIONAL_SAMPLING_CHUNK
   GWForge.population.mass.NODES_PER_PEAK_WIDTH
   GWForge.population.mass.MINIMUM_MASS_GRID_NODES
   GWForge.population.mass.MAXIMUM_MASS_GRID_NODES


Classes
-------

.. autoapisummary::

   GWForge.population.mass.Mass


Functions
---------

.. autoapisummary::

   GWForge.population.mass.notch_filter
   GWForge.population.mass.low_pass_filter
   GWForge.population.mass.primary_mass_grid_nodes
   GWForge.population.mass.sample_mass_ratio_given_primary


Module Contents
---------------

.. py:function:: notch_filter(val, parameters)

.. py:function:: low_pass_filter(val, parameters)

.. py:data:: choices
   :value: ['PowerLaw+Peak', 'MultiPeak', 'BrokenPowerLaw', 'BGP', 'UniformSecondary', 'DoubleGaussian',...


.. py:data:: sampler_choices
   :value: ['importance_m1_m2', 'importance_m1_q', 'lint']


.. py:data:: BGP_PARAMETERS

.. py:data:: POWER_LAW_PEAK_PARAMETERS

.. py:data:: DEFAULT_PARAMETERS

.. py:data:: CONDITIONAL_SAMPLING_CHUNK
   :value: 4096


.. py:data:: NODES_PER_PEAK_WIDTH
   :value: 40


.. py:data:: MINIMUM_MASS_GRID_NODES
   :value: 1000


.. py:data:: MAXIMUM_MASS_GRID_NODES
   :value: 40000


.. py:function:: primary_mass_grid_nodes(parameters, minimum_mass, maximum_mass)

   Number of primary-mass grid nodes needed to resolve the narrowest peak.

   The smoothed mass models are sampled by wrapping the density in a
   :class:`bilby.core.prior.Interped`, which builds a **linear interpolant** on
   this grid and inverse-CDF samples from that. The interpolant, not the
   density, is what the catalogue is drawn from, so the grid has to resolve
   every feature the density has.

   :param parameters: Mass-model parameters. Any key beginning with ``sigpp`` is treated as a
                      peak width; models without peaks keep the floor.
   :type parameters: dict
   :param minimum_mass: Range the grid spans.
   :type minimum_mass: float
   :param maximum_mass: Range the grid spans.
   :type maximum_mass: float

   :returns: Clipped to ``[MINIMUM_MASS_GRID_NODES, MAXIMUM_MASS_GRID_NODES]``.
   :rtype: int


.. py:function:: sample_mass_ratio_given_primary(model, primary_samples, mass_ratio_parameters, random_state=None)

   Draw ``mass_ratio`` conditionally on ``mass_1``, by inverse CDF.

   :param model: Supplies the ``m1s``/``qs`` grids and ``p_q``.
   :type model: BaseSmoothedMassDistribution
   :param primary_samples: Sampled ``mass_1_source``.
   :type primary_samples: array_like
   :param mass_ratio_parameters: ``beta``, ``mmin`` and ``delta_m``, and optionally ``mmin_2`` and
                                 ``delta_m_2`` for models whose secondary has its own taper.
   :type mass_ratio_parameters: dict
   :param random_state: Defaults to numpy's global RNG, so a single seed governs it.
   :type random_state: numpy.random.Generator or None

   :returns: Mass-ratio samples, the same length as ``primary_samples``.
   :rtype: numpy.ndarray


.. py:class:: Mass(mass_model='BGP', number_of_samples=1, parameters=None, full_pop_sampler='importance_m1_m2')

   .. py:attribute:: mass_model
      :value: ''



   .. py:attribute:: number_of_samples
      :value: 1



   .. py:attribute:: parameters


   .. py:attribute:: full_pop_sampler
      :value: 'importance_m1_m2'



   .. py:method:: sample(**sampler_kwargs)

      Generate mass distribution samples based on the chosen parameterised model and its parameters.

      :returns: **dict**
      :rtype: A dictionary containing source frame mass distribution samples.