"""
Module containing the `~halotools.empirical_models.SubhaloMockFactory` class,
the primary class used to construct mock galaxy populations
based on subhalo-based models.
The `~halotools.empirical_models.SubhaloMockFactory` class
provides an abstract interface between halo catalogs
and Halotools models.
"""
import numpy as np
from astropy.table import Table
from .mock_factory_template import MockFactory
from .. import model_defaults
from ...custom_exceptions import HalotoolsError
__all__ = ['SubhaloMockFactory']
__author__ = ['Andrew Hearin']
unavailable_haloprop_msg = ("Your model requires that the ``%s`` key appear in the halo catalog,\n"
"but this column is not available in the catalog you attempted to populate.\n")
[docs]
class SubhaloMockFactory(MockFactory):
""" Class responsible for populating a simulation with a
population of mock galaxies based on models generated by
`~halotools.empirical_models.SubhaloModelFactory`.
Can be thought of as a factory that takes a model
and simulation halocat as input,
and generates a mock galaxy population.
The returned collection of galaxies possesses whatever
attributes were requested by the model, such as xyz position,
central/satellite designation, star-formation rate, etc.
See :ref:`subhalo_mock_factory_source_notes`
for an in-depth tutorial on the mock-making algorithm.
"""
def __init__(self, **kwargs):
"""
Parameters
----------
halocat : object, keyword argument
Object containing the halo catalog and other associated data.
Produced by `~halotools.sim_manager.CachedHaloCatalog`
model : object, keyword argument
A model built by a sub-class of `~halotools.empirical_models.SubhaloModelFactory`.
populate : boolean, optional
If set to ``False``, the class will perform all pre-processing tasks
but will not call the ``model`` to populate the ``galaxy_table``
with mock galaxies and their observable properties. Default is ``True``.
"""
MockFactory.__init__(self, **kwargs)
halocat = kwargs['halocat']
# Pre-compute any additional halo properties required by the model
self.preprocess_halo_catalog(halocat)
self.precompute_galprops()
[docs]
def preprocess_halo_catalog(self, halocat):
""" Method to pre-process a halo catalog upon instantiation of the mock object.
"""
halo_table = halocat.halo_table
if (('halo_hostid' not in self.additional_haloprops) & ('halo_hostid' in list(halo_table.keys()))):
self.additional_haloprops.append('halo_hostid')
if (('halo_mvir_host_halo' not in self.additional_haloprops) &
('halo_mvir_host_halo' in list(halo_table.keys()))):
self.additional_haloprops.append('halo_mvir_host_halo')
self.halo_table = Table()
for key in self.additional_haloprops:
try:
self.halo_table[key] = halo_table[key]
except KeyError:
raise HalotoolsError(unavailable_haloprop_msg % key)
[docs]
def precompute_galprops(self):
""" Method pre-processes the input subhalo catalog, and pre-computes
all halo properties that will be inherited by the ``galaxy_table``.
For example, in subhalo-based models, the phase space coordinates of the
galaxies are hard-wired to be equal to the phase space coordinates of the
parent subhalos, so these keys of the galaxy_table
can be pre-computed once and for all.
Additionally, a feature of some composite models may have explicit dependence
upon the type of halo/galaxy. The `gal_type_func` mechanism addresses this potential need
by adding an additional column(s) to the galaxy_table. These additional columns
can also be pre-computed as halo types do not depend upon model parameter values.
"""
self._precomputed_galprop_list = []
for key in self.additional_haloprops:
self.galaxy_table[key] = self.halo_table[key]
self._precomputed_galprop_list.append(key)
phase_space_keys = ['x', 'y', 'z', 'vx', 'vy', 'vz']
for newkey in phase_space_keys:
self.galaxy_table[newkey] = self.galaxy_table[model_defaults.host_haloprop_prefix+newkey]
self._precomputed_galprop_list.append(newkey)
self.galaxy_table['galid'] = np.arange(len(self.galaxy_table))
self._precomputed_galprop_list.append('galid')
# Component models may explicitly distinguish between certain types of halos,
# e.g., subhalos vs. host halos. Since this assignment is not dynamic,
# it can be pre-computed.
for feature, component_model in self.model.model_dictionary.items():
try:
f = component_model.gal_type_func
newkey = feature + '_gal_type'
self.galaxy_table[newkey] = f(table=self.galaxy_table)
self._precomputed_galprop_list.append(newkey)
except AttributeError:
pass
except:
clname = component_model.__class__.__name__
msg = ("\nThe `gal_type_func` attribute of the " + clname +
"\nraises an unexpected exception when passed a halo table as a "
"table keyword argument. \n"
"If the features in your component model have explicit dependence "
"on galaxy type, \nthen you must implement the `gal_type_func` mechanism "
"in such a way that\nthis function accepts a "
"length-N halo table as a ``table`` keyword argument, \n"
"and returns a length-N array of strings.\n")
raise HalotoolsError(msg)
[docs]
def populate(self, seed=None):
"""
Method populating subhalos with mock galaxies.
By calling the `populate` method of your mock, you will repopulate
the halo catalog with a new realization of the model based on
whatever values of the model parameters are currently stored in the
``param_dict`` of the model.
For an in-depth discussion of how this method is implemented,
see the :ref:`subhalo_mock_factory_source_notes` section of the documentation.
Parameters
----------
seed : int, optional
Random number seed used in the Monte Carlo realization.
Default is None, which will produce stochastic results.
Notes
-----
Note the difference between the
`halotools.empirical_models.SubhaloMockFactory.populate` method and the
closely related method
`halotools.empirical_models.SubhaloModelFactory.populate_mock`.
The `~halotools.empirical_models.SubhaloModelFactory.populate_mock` method
is bound to a composite model instance and is called the *first* time
a composite model is used to generate a mock. Calling the
`~halotools.empirical_models.SubhaloModelFactory.populate_mock` method creates
the `~halotools.empirical_models.SubhaloMockFactory` instance and binds it to
composite model. From then on, if you want to *repopulate* a new Universe
with the same composite model, you should instead call the
`~halotools.empirical_models.SubhaloMockFactory.populate` method
bound to ``model.mock``. The reason for this distinction is that
calling `~halotools.empirical_models.SubhaloModelFactory.populate_mock`
triggers a large number of relatively expensive pre-processing steps
and self-consistency checks that need only be carried out once.
See the Examples section below for an explicit demonstration.
In particular, if you are running an MCMC type analysis,
you will choose your halo catalog and completeness cuts, and call
`~halotools.empirical_models.SubhaloModelFactory.populate_mock`
with the appropriate arguments. Thereafter, you can
explore parameter space by changing the values stored in the
``param_dict`` dictionary attached to the model, and then calling the
`~halotools.empirical_models.MockFactory.populate` method
bound to ``model.mock``. Any changes to the ``param_dict`` of the
model will automatically propagate into the behavior of
the `~halotools.empirical_models.MockFactory.populate` method.
Normally, repeated calls to the
`~halotools.empirical_models.SubhaloMockFactory.populate` method
should not increase the RAM usage of halotools because a new mock
catalog is created and the old one
deleted. However, on certain machines the memory usage was found to
increase over time. If this is the case and memory usage is critical you
can try calling gc.collect() immediately following the call to
``mock.populate`` to manually invoke python's garbage collection.
Examples
----------
>>> from halotools.empirical_models import PrebuiltSubhaloModelFactory
>>> model_instance = PrebuiltSubhaloModelFactory('behroozi10')
Here we will use a fake simulation, but you can populate mocks
using any instance of `~halotools.sim_manager.CachedHaloCatalog` or
`~halotools.sim_manager.UserSuppliedHaloCatalog`.
>>> from halotools.sim_manager import FakeSim
>>> halocat = FakeSim()
>>> model_instance.populate_mock(halocat)
Your ``model_instance`` now has a ``mock`` attribute bound to it.
You can call the `populate` method bound to the ``mock``,
which will repopulate the halo catalog with a new Monte Carlo
realization of the model.
>>> model_instance.mock.populate()
If you want to change the behavior of your model, just change the
values stored in the ``param_dict``. Differences in the parameter values
will change the behavior of the mock-population.
>>> model_instance.param_dict['scatter_model_param1'] = 0.25
>>> model_instance.mock.populate()
See also
---------
:ref:`subhalo_mock_factory_source_notes`
"""
self._allocate_memory(seed=seed)
for method in self.model._mock_generation_calling_sequence:
func = getattr(self.model, method)
if seed is not None:
seed += 1
func(table=self.galaxy_table, seed=seed)
if hasattr(self.model, 'galaxy_selection_func'):
mask = self.model.galaxy_selection_func(self.galaxy_table)
self.galaxy_table = self.galaxy_table[mask]
def _allocate_memory(self, seed=None):
"""
"""
Ngals = len(self.galaxy_table)
new_column_generator = (key for key in self.model._galprop_dtypes_to_allocate.names
if key not in self._precomputed_galprop_list)
for key in new_column_generator:
dt = self.model._galprop_dtypes_to_allocate[key]
self.galaxy_table[key] = np.empty(Ngals, dtype=dt)