Halotools is a specialized python package for building and testing models of the galaxy-halo connection, and analyzing catalogs of dark matter halos. The core feature of Halotools is a modular platform for creating mock universes of galaxies starting from a catalog of dark matter halos obtained from a cosmological simulation. Functionality of the package includes:
- Fast generation of synthetic galaxy populations using HODs, abundance matching, and related methods
- Efficient algorithms for calculating galaxy clustering, lensing, z-space distortions, and other astronomical statistics
- A modular, object-oriented framework for designing your own galaxy evolution model
- End-to-end support for downloading publicly-available halo catalogs and reducing them to fast-loading hdf5 files
The code is publicly available at https://github.com/astropy/halotools.
The simplest and most reliable way to install the latest release of the code is with conda-forge:
conda install -c conda-forge halotools
Pip installation is not recommended because the conda-forge dependency solver is much more robust. However, users who prefer pip can install via:
pip install halotools
You can find detailed installation instructions halotools.readthedocs.io.
After installing the package, you should navigate to the Quickstart Guides and Tutorials section and follow the Getting started with Halotools 10-minute tutorial. This will get you set up with the default halo catalog so that you can quickly get started with creating mock galaxy populations.
The latest build of the documentation can be found at http://halotools.readthedocs.io. The documentation includes installation instructions, quickstart guides and step-by-step tutorials. The Basic features section below gives an overview of the primary functionality of the package.
Once you have installed the code and downloaded the default halo catalog (see the Getting Started guide in the documentation), you can use Halotools models to populate mock galaxy populations.
# Select a model
from halotools.empirical_models import PrebuiltHodModelFactory
model = PrebuiltHodModelFactory('zheng07')
# Select a halo catalog
from halotools.sim_manager import CachedHaloCatalog
halocat = CachedHaloCatalog(simname='bolshoi', redshift=0, halo_finder='rockstar')
# populate the catalog with the model
model.populate_mock(halocat)After calling populate_mock, your model will have a mock attribute storing your synthetic galaxy population. All Halotools models have a populate_mock method that works in this way, regardless of the features of the model. There are no restrictions on the simulation or halo-finder with which you can use the populate_mock method.
All Halotools models have a param_dict that controls the behavior of the model. By changing the parameters in this dictionary, you can create alternative versions of your mock universe by re-populating the halo catalog as follows.
model.param_dict['logMmin'] = 12.1
model.mock.populate()Note how much faster the call to mock.populate is relative to model.populate_mock(halocat). This is due to a large amount of one-time-only pre-processing that is carried out upon creation of the first mock universe. The process of varying param_dict values and repeatedly calling model.mock.populate() is part of a typical workflow in an MCMC-type analysis conducted with Halotools.
The pre-built model factories give you a wide range of models to choose from, each based on an existing publication. Alternatively, you can use the Halotools factories to design a customized model of your own creation, such as models for stellar mass, color, size, morphology, or any property of your choosing. The modular design of the empirical_models sub-package allows you to mix-and-match an arbitrary number or kind of features to create your own composite model of the full galaxy population. You can choose from component models provided by Halotools, components exclusively written by you, or anywhere in between. Whatever science features you choose, any Halotools model can populate any Halotools-formatted halo catalog with the same syntax shown above.
The mock_observables sub-package contains a wide variety of optimized functions that you can use to study your mock galaxy population. For example, you can calculate projected clustering via the wp function, identify friends-of-friends groups with FoFGroups, or compute galaxy-galaxy lensing with mean_delta_sigma.
from halotools.mock_observables import wp
from halotools.mock_observables import FoFGroups
from halotools.mock_observables import mean_delta_sigmaThere are many other functions provided by the mock_observables package, such as RSD multipoles, pairwise velocities, generalized marked correlation functions, customizable isolation criteria, void statistics, and more.
Halotools provides end-to-end support for downloading simulation data, reducing it to a fast-loading hdf5 file with metadata to help with the bookkeeping, and creating a persistent memory of where your data is stored on disk. This functionality is handled by the sim_manager sub-package:
from halotools import sim_managerThe sim_manager package comes with a memory-efficient TabularAsciiReader designed to handle the very large file sizes that are typical of contemporary cosmological simulations. There are 20 halo catalogs available for download from the Halotools website using the download_additional_halocat script.py, including simulations run with different volumes, resolutions and cosmologies, and also catalogs identified using different halo-finders and at different redshift. Any simulation you store in cache can be loaded into memory in the same way, and all such catalogs have a halo_table attribute storing the actual data.
from halotools.sim_manager import CachedHaloCatalog
halocat = CachedHaloCatalog(simname=any_simname, redshift=any_redshift, halo_finder=any_halo_finder)
print(halocat.halo_table[0:10])You are not limited to use the halo catalogs pre-processed by Halotools. The UserSuppliedHaloCatalog allows you to use your own simulation data and transform it into a Halotools-formatted catalog in a simple way.
from halotools.sim_manager import UserSuppliedHaloCatalogAlthough the sim_manager provides an object-oriented framework for creating a persistent memory of where you store your halo catalogs, your cache is stored in a simple, human-readable ASCII log in the following location:
$HOME/.astropy/cache/halotools/halo_table_cache_log.txt
Halotools is a fully open-source package with contributing scientists spread across many universities. The latest stable release of the package, v0.9, is now available on pip and conda-forge. You can also install the development version of the package by cloning the master branch on GitHub and locally building the source code, as described in the installation instructions.
## Asking questions and staying up-to-date
You can contact Andrew Hearin directly by email at ahearin-at-anl-dot-gov, or by tagging @aphearin on GitHub.
If you use Halotools modules to support your science publication, please cite Hearin et al. (2017), ideally taking note of the version of the code you used, e.g., v0.8:
@ARTICLE{halotools,
author = {{Hearin}, Andrew P. and {Campbell}, Duncan and {Tollerud}, Erik and {Behroozi}, Peter and {Diemer}, Benedikt and {Goldbaum}, Nathan J. and {Jennings}, Elise and {Leauthaud}, Alexie and {Mao}, Yao-Yuan and {More}, Surhud and {Parejko}, John and {Sinha}, Manodeep and {Sip{\"o}cz}, Brigitta and {Zentner}, Andrew},
title = "{Forward Modeling of Large-scale Structure: An Open-source Approach with Halotools}",
journal = {The Astronomical Journal},
keywords = {cosmology: theory, galaxies: halos, galaxies: statistics, large-scale structure of universe, Astrophysics - Instrumentation and Methods for Astrophysics, Astrophysics - Cosmology and Nongalactic Astrophysics, Astrophysics - Astrophysics of Galaxies},
year = 2017,
month = nov,
volume = {154},
number = {5},
eid = {190},
pages = {190},
doi = {10.3847/1538-3881/aa859f},
archivePrefix = {arXiv},
eprint = {1606.04106},
primaryClass = {astro-ph.IM},
adsurl = {https://ui.adsabs.harvard.edu/abs/2017AJ....154..190H},
adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}
Halotools is licensed by Argonne National Lab under a 3-clause BSD style license - see the licenses/LICENSE.rst file.