search

transit-timing / tt

TESS Transit Timing of 278 Planets
0 stars 1 forks source link

readme

TESS Transit Timing of Hundreds of Hot Jupiters DOI

This repository contains source code for the "TESS Transit Timing of Hundreds of Hot Jupiters" paper. The repository provides code (1) to process TESS light curves (with 2, 10, or 30-minute sampling) to extract transit times, and (2) to automatically download arXiv articles that contain a given planet's name and a number greater than 2,000,000. A detailed description of the code can be found in the paper.

Processing TESS light curves

The code relies on the following Python packages: numpy, pandas, scikit-learn, astropy, matplotlib, batman-package, scipy, lightkurve, lmfit, lxml, arxiv. You may install the packages via pip or, if you use anaconda, you may run the following commands:

module load anaconda
conda create -n tt python=3.7 numpy pandas scikit-learn astropy matplotlib batman-package scipy lightkurve lmfit lxml arxiv
conda activate tt

Then clone this repository to your machine:

git clone https://github.com/transit-timing/tt.git

To process TESS light curves stored in the ~/data/ folder and extract mid-transit times, first go to the directory containing the code:

cd 0_tt

Inside main.py script, you may specify the name of the planet whose TESS light curve you would like to process. Then, you can run the script:

python3 main.py

Output

The results are stored in the ~/output folder. A typical folder that this script produces for a given target contains the following:

  1. Figure of the TESS light curve that was processed (transits that were processed to extract mid-transit times are shown in red):
drawing
  1. Figure with the individual transits that were processed (best-fit transit model is shown in red):
drawing
  1. Folded light curve:
drawing
  1. O-C diagram showing timing residuals. The timing residuals were calculated as follows: first, a linear model was fitted to the transit times extracted from TESS data. Then, the linear model was subtracted from the transit times at each epoch:
drawing
  1. planet_name_results.txt contains extracted mid-transit times and their uncertainties.
  2. planet_name_log.txt contains intermediate outputs of the code, such as the found best-fit transit model parameters and different statistics of the produced fits.

Note: in order to process TESS light curves, you may want to first download the relevant .fits files. To download a given light curve, you may do the following:

cd 0_download_data

In the download_single_target.py script, specify the name of the target and the light curve that you would like to download. Then run the script:

python3 download_single_target.py

Figures

O-C diagrams for each of the 382 systems are available here.

Tables

  1. A single CSV file with all of the collected transit times for each target is available at https://github.com/transit-timing/transit-timing/blob/main/3_database/table4.csv

  2. A single CSV file with all of the target ephemerides is available at https://github.com/transit-timing/transit-timing/blob/main/3_database/table3.csv

    Scraping literature

    A semi-automated procedure was used to search the extensive literature on transiting planets to find all the previously reported times for all the systems in our sample. The details of our approach to scraping the data can be found in Section 4 of the paper. The basic steps are as follows.

  3. Download bibcodes of all papers related to a given object from the NASA Astrophysics Data System database:

    • Go to https://ui.adsabs.harvard.edu/search;
    • Do cone search using RaDec coordinates with 3" radius. For example, you may execute the following search: 
      object:"20h30m54.13s 6d25m46.4s:3"

      Alternatively, you can type object:"planet_name" in the search bar of the NASA Astrophysics Data System database to retrieve articles for a planet with a given name planet_name. For example, you may type:

      object:"WASP-12"

      Then click "Export" -> "in BibTeX" and download the .bib file containing all of the papers referencing a given planet. Place the downloaded .bib file into ~/arxiv/article_database folder and rename the file to whatever the name of the planet is (e.g. WASP-12.bib)

  4. Run bibcode2arxiv_id.py to create a .txt file containing a list of arXiv IDs of papers related to a given object whose name is planet_name. The arXiv IDs are extracted from the .bib file downloaded in the previous step:

    cd 1_scrape_literature
    python3 bibcode2arxiv_id.py -p planet_name

    For example, if you downloaded a .bib file for WASP-12, then you would run

    python3 bibcode2arxiv_id.py -p WASP-12

  5. Execute run.py to download .tex source files of arXiv papers retrieved in the previous step. The script untars the source files and transfer all .tex files to a different directory.

    python3 -u run.py -p WASP-12

    For a given planet, all tar.gz are stored in ~/arxiv_dir/{planet_name} directory. For a given planet, all untarred files are stored in ~/untar_dir/{planet_name} directory.

  6. Run read_tex.py to find numbers > 2,000,000; if such numbers and the name of the planet are present in a given paper, a PDF of the paper is transferred to ~/{planet_name} directory for a final manual review.

python3 read_tex.py -s wasp -id 12

Citation

If you find the code useful, consider citing it: DOI