Welcome to enterprise_warp’s documentation!
enterprise_warp
enterprise_warp is a set of tools for pulsar timing analysis with Enterprise, Libstempo and Tempo2. It uses the Bilby to enable Bayesian inference for pulsar timing arrays using many different samplers and all other advantages of Bilby.
Look how easy it is to run from the command line:
$ python run_example_paramfile.py --prfile example_params/default_model_dynesty.dat --num 0
Here run_example_paramfile.py is the script available in examples/. To view and analyze the results, run:
$ python -m enterprise_warp.results --result example_params/default_model_dynesty.dat --info 1 --corner 1
Where --result can be a parameter file or an output directory. The above command saves a corner plot to the output directory. Other command line options are related to noise files, Bayes factors, chain plots, etc. Please run python -m enterprise_warp.results -h to list available options.
Features
Configure your runs with parameter files instead of editing python scripts. This allows easier management of your data analysis progress. Keep track of your input and output directories.
The code is optimized for parallel computing via the usage of command line arguments (i.e., for different pulsars)
Simply choose any MCMC/Nested sampler that is available in Bilby: PTMCMCSampler, Dynesty, Nestle, PyPolyChord, PyMC3, ptmcee, and more. Evaluate evidence, perform parameter estimation, or directly compare models with product-space method and PTMCMCSampler.
Accellerate your data analysis with MPI using PyPolychord sampler: distribute heavy calculations between ~10-100 cores of a supercomputer. The code has special options for MPI runs.
Use noise model files with tailored noise models for each pulsar and common noise processes in all pulsars. Run analysis with individual pulsars or the whole pulsar timing array.
Save time on setting up runs: all sampler keyword arguments, your custom noise models and priors are automatically recongized in parameter files.
Easily add your own models, using several examples. Or load models from enterprise_extensions or other code.
And more!
Getting started
You should simply have the main running script which imports enterprise_warp. When running it, you should point it to a parameter file. If necessary, you can also have a custom noise model file .py with a parent class StandardModels, with custom models for signal and noise. It is then imported in the main running script.
When starting your new project with enterprise_warp, you can copy/fork examples below:
PPTA DR2 noise analysis (2019-2020). Here, the main running script is
run_dr2.pyand custom noise models are inppta_dr2_models.py.Search for the gravitational wave background with PPTA DR2 (2020-2021). Here, the main running script is
run_analysis.pyand custom noise models are inppta_dr2_models.py.
License
The project is licensed under the MIT license.