pyorbit-package 9.0.22

PyORBIT: a code for exoplanet orbital parameters and stellar activity

A code for exoplanet orbital parameters and stellar activity.

---

PyORBIT version 9.0.21 by Luca Malavolta - December 2022

News

• PyORBIT will use the correct parametrization when a prior is assigned to a parameter and nested sampling is used (NS do not allow to assign priors to derived parameters).

• Many new models are now available

  • Multidimensional Gaussian Process (also known as GP Framework, Rajpaul et al. 2015, Barragán et al. 2022).
  • New kernels for Gaussian Process regression: Quasi-Periodic with Cosine (Perger et al. 2021), Quasi-Periodic with first derivative.
  • CHEOPS detrending (similar to FactorModel as in pycheops).
  • Phase curve and secondary eclipse modelling with batman (Kreidberg 2015) or spiderman (Louden et al. 2016).
  • Rossiter-McLaughlin through analytical formulation by Ohta et al. 2005.
  • Celerite/Celerite2 standard models, now better organized.
  • Fit of individual time of transits (with automatic definition of boundaries) for TTV analysis.
  • experimenting with tinygp

• More options for parameters space exploration: Linear, Log_Natural, Log_Base2, Log_Base10.

• Some variables have been renamed, to improve clarity of results:

  Definition	PyORBIT 8.x	PyORBIT 9.x
  Mean Longitude	f	mean_long
  Scaled planetary radius	R	R_Rs
  Scaled semi-major axis	a	a_Rs
  Planetary mass in Earth masses	M	M_Me
  Stellar density	rho	density

  Also, batman_ld_quadratic and pytransit_ld_quadratic have been merged into ld_quadratic.

  Note: the Mean Longitude is defined assuming the longitude of the ascending node $\Omega$ equal to zero, thus corresponding to the angle defined in section 4.3 of (Ford 2006) and simply called phase in Malavolta et al. (2016)

• Overall reorganization of the code

Warning Loss of backward-compatibility

You cannot analyzes results obtain with previous versions (< 9) of PyORBIT. No worries, the old version is still available in the legacy branch, you can download it from the Github page or switch to it through the terminal:

git checkout legacy

To switch back to the current version, just execute:

git checkout main

Working on it

• Rossiter-McLaughlin through starry (Luger et al. 2019)
• Multi-component GP for light curves (Rotation + Granulation, Rotation + Granulation + Oscillations), following Barros et al. 2020.

Samplers

Bayesian evidence estimation can now be performed with:

• dynesty by Josh Speagle dynesty documentation, code repository
• UltraNest by Johannes Buchner UltraNest documentation, code repository
• Multinest and Polychord support is still there, but not supported anymore

Just substitute "emcee" with "dynesty" or "ultranest" to when running the code. Warning MCMC and Nested Sampling handle prior in a radically different way, as such it s not possible to directly translate some priors from one sampler to another

Documentation Some incomplete documentation is available here. For any doubt, feel free to contact me at luca.malavolta_at_unipd.it, I'll be happy to work out together any problem that may arise during installation or usage of this software.

PyORBIT handles several kinds of datasets, such as radial velocity (RV), activity indexes, and photometry, to simultaneously characterize the orbital parameters of exoplanets and the noise induced by the activity of the host star. RV computation is performed using either non-interacting Kepler orbits or n-body integration. Stellar activity can be modeled either with sinusoids at the rotational period and its harmonics or gaussian process. Offsets and systematics in measurements from several instruments can be modeled as well. Thanks to the modular approach, new methods for stellar activity modeling or parameter estimation can be easily incorporated into the code.

Models Any of these models can be applied to a dataset. The user can choose which models should be used for each dataset.

• Gaussian Processes for RV or photometry (shared or independent hyperparameter)
• Transits, eclipses, phase curves
• Polynomial trends with user-defined order
• Correlation with activity indexes (or any other dataset)
• Sinusoids (independent or shared amplitudes and periods)
• Harmonics to test old results in the literature

Priors These priors can be applied to any of the parameters (it's up to the user to choice the appropriate ones):

• Uniform default prior for all the parameters
• Gaussian
• Jeffreys
• Modified Jeffreys
• Truncated Rayleigh
• WhiteNoisePrior
• BetaDistribution

Jeffreys and Modified Jeffreys priors are actually Truncated Jeffreys and Truncated Modified Jeffreys, with truncation defined by the boundaries of the parameter space.

Parameter exploration The user can choice between Linear and Logarithmic. Note that in the second case the parameter space is transformed into base-2 logarithm.

Most of the information can be found in Malavolta et al. (2016) and Malavolta et al. (2018).