Tutorials

Preparing data

The functions in exomagpy.predictExo take in data in the form of a .csv file containing a list of target names. For TESS targets, the TIC IDs are used. For Kepler targets, KIC IDs are used. predictExo will read the contents of any column with the header “tic_id”, “tid”, “kic-id” or “kid”.

Predicting exoplanets

exomagpy can predict the existence of exoplanets around a target star for most TESS and Kepler targets. To predict exoplanet candidates, you can use the exomagpy.predictExo.tess() or exomagpy.predictExo.kepler() functions as below:

import exomagpy.predictExo
exomagpy.predictExo.tess(train1,size1,train2,size2,test,testsize)
exomagpy.predictExo.kepler(train1,size1,train2,size2,test,testsize)

Here, train1 is a training dataset containing targets with confirmed exoplanets, and train2 is a training dataset containing targets with confirmed no exoplanets. size1 and size2 are the number of targets you wish to use from these training datasets respectively. test is the test dataset (the data you want to make predictions for) and testsize is the number of targets to use for that dataset.

When the functions are run, the program will search for the specified targets using Lightkurve. After the targets are found, the function’s neural network will train using the lightcurves of the given train data. After training, the program will output the plot for each lightcurve (if running in Jupyter notebook) alongside its predicted exoplanetary status and ID.

Downloading lightcurves

exomagpy has a function lc_to_array, which finds the lightcurve for an individual target and converts it into an 8-bit numpy image array.

import exomagpy.download
exomagpy.download.lc_to_array(search)

where search is a LightCurve object.