hposborn / monotools

A package for detecting, vetting, and modelling transiting exoplanets on uncertain periods

MonoTools: A python package for planets of uncertain period

MonoTools is a package specific to the detection, vetting and modelling of transiting exoplanets, with a specific emphasis on Monotransiting planets, and those with unknown periods.

MonoSearch includes scripts specifically for searching and assessing a lightcurve for the presence of monotransits. This includes:

• Searching for monotransits
• Searching for periodic planets
• Performing a best-fit transit model
• Vetting whether detected monotransits are variability, asteroids (using background flux), background EB (with centroids), etc.
• Assessing whether those transits are linked to any detected multi-transiting planet candidate, or with each other - e.g. a 2-transit duo.
• Setting up the monotransit fitting code.
• Fitting planets in a Bayesian way to account for uncertain periods (i.e. mono or duo-transits), lightcurve gaps, stellar variability, etc.
• Compiling all the steps and various plots into a report for each planet candidate

Full documentation is are taking shape at readthedocs.

MonoTools.fit is an update to my previous Namaste code using the "exoplanet" models of Dan Foreman-Mackey et al.

This uses a PyMC3/Theano back-end, a differentiable Keplerian orbital model, and the Gaussian process modeling of Celerite to fit exoplanetary transits.

Here we wrap the "exoplanet" core model to specify it for single and "duo"-transits, and plot/analyse the results.

The first use of MonoTools in a published paper is out now in Osborn+ (2022) where MonoTools helped recover the orbits of the two outer planets in the TOI-2076 system.

Full installation and usage advice on "ReadTheDocs"

Installing

To install, I recommend using a virtual environment, as some of the packages required are not at their most recent versions. MonoTools should be pip installable, therefore run pip install MonoTools.

Alternatively, to run the most up-to-date development version, you can run git clone http://github.com/hposborn/MonoTools, cd into the MonoTools folder, then run pip install . (plus make sure the folder where MonoTools is installed is included in your $PYTHONPATH, e.g. by adding export PYTHONPATH=/path/to/dir:$PYTHONPATH to your .bashrc file).

MonoTools will look at the $MONOTOOLSPATH bash variable as the location to store files, and defaults to MonoTools/data. Include this in your .bashrc file to modify this location.

Mac OSX users may need to make sure GCC is correctly installed (.e.g with brew install gcc) and C libraries are present.

Using MonoTools

For a tutorial on how to deal with the in-built lightcurve class, see using_lightcurve.ipynb.

For examples how to run the tools (e.g. lightcurve, stellar parameters, etc) and the search (monotransit search and vetting) functions, see using_fit.ipynb [TBD]. https://github.com/hposborn/MonoTools/blob/master/MonoTools/examples/Search_Example.ipynb).

For examples on how to run the fitting code (including with RVs), see using_fit.ipynb.

<--

Extra steps to install:

If stellar parameters are not given and not accessible from the input catalogues provided, Dan Huber's "Isoclassify" is required to estimate density given info scraped in the stellar module from Gaia, input catalogues, wide-field surveys, spectra, etc. A modified version is included. However, the "mesa.h5" file must be downloaded using wget https://www.dropbox.com/s/vrr8hc7qav1fzyb/mesa.h5?dl=0 in stellar/isochrones. mwdust modules may also require specific installation.-->

monotools's Issues

Pip install problem with batman-package 2.4.7 (windows)

Hi,

I tried to pip install monotools in a fresh windows env and batman-package 2,4,7 throws an error (same when I try to install it alone)
But I can install batman 2.4.8 without problems.
Maybe you can upgrade batman version in the requirements?

Thank you!
Rafael