EMPEROR Exoplanet Mcmc Parallel tEmpering Radial velOcity fitteR

EMPEROR (Exoplanet Mcmc Parallel tEmpering Radial velOcity fitteR), is a new Python-based algorithm that automatically searches for signals in radial velocity timeseries, employing Markov chains and parallel tempering methods, convergence tests and Bayesian statistics, along with various noise models. A number of posterior sampling routines are available, focused on efficiently searching for signals in highly multi-modal posteriors. The code allows the analysis of multi-instrument and multi-planet data sets and performs model comparisons automatically to return the optimum model that best describes the data.

This code makes use of:

• Numpy
• Scipy
• emcee (http://dan.iel.fm/emcee/current/)
• PyAstronomy (http://www.hs.uni-hamburg.de/DE/Ins/Per/Czesla/PyA/PyA/index.html)
• pygame (https://www.pygame.org/)
• tqdm (https://pypi.python.org/pypi/tqdm)
• termcolor (https://pypi.python.org/pypi/termcolor)
• corner (https://pypi.python.org/pypi/corner)
• reddutils (https://github.com/ReddTea/reddutils)
• tabulate (https://pypi.org/project/tabulate/)
• kepler (https://github.com/dfm/kepler.py)
• arviz (https://arviz-devs.github.io/arviz/)

All of them can be easily installed with pip.

Just try out

pip3 install astroEMPEROR

For the 'manual installation' read below!!

In the console type in your work folder

git clone https://github.com/ReddTea/astroEMPEROR.git

Download the tests folder and run emp_test.py to make sure everything works!

ipython  # open python environment
run emp_test

or just

python emp_test.py

The code is really simple to use! You just need to have your data under /datafiles folder. And then, as shown in python_file.py, append the names of the datasets you want to use in an array, set the configuration of the chain in another single array, then call the method and conquer!

import astroEMPEROR as emp
import numpy as np

sim = emp.Simulation()
sim.set_engine('emcee')
setup = np.array([3, 50, 250, 2])
sim.load_data('51Peg')  # Target folder name in /datafiles/
sim.run_auto(setup, k_start=0, k_end=2)

They go under /datalogs folder. Following the name convention for the datasets, you should have them properly classified as /datalogs/starname/<date_i>, where i is the ith run in that date.

You will see chain plots, posterior plots, histograms, phasefolded curves, the chain sample and more!!!

• It's really simple to use
• It has a series of configuration commands that will amaze you
• Advanced Noise Model
• Quite Flexible!

I'll update this soon, promise, in the meanwhile, there is a reference guide at the bottom. See the example file for the know-how.

run_auto(setup, k_start=2, k_end=2, parameterisation=int, moav=int, accel=int)

setup: Sets the number of temperatures, walkers and steps you will use, respectively

up_to_k: Up to the k-th keplerian signal

parameterisation: Sets the parameterisation to use for the keplerian model.

acc: Sets the acceleration order. 0 for no acceleration, 1 for linear acceleration, 2 for linear plus quadratic...

moav: Sets the moving average order.

.add_condition([param_name, attribute, value])

This modifies the corresponding parameter. It's equivalent to do in the proper part of the run: Parameter[param_name].attribute = value

param_name: str with the name of the parameter.

attribute: str with the name of any attribute of the Parameter object.

value: The value you want to change it to.