Code for a quasar variability study in collaboration with @esellers04
This code is designed to complete a quasar variability study using Transiting Exoplanet Survey Satellite (TESS) data and Long Short-Term Memory Recurrent Neural Networks (LSTM-RNN).
To operate effectively, clone to a repository and set up a a second directory where data can be stored. Run quasar_catalogue.csv with data from sources below to obtain the catalogue and its contents or email alexander.evitt@gmail.com for a copy of the data.
Catalogue of quasars taken from https://www.sdss.org/dr12/algorithms/boss-dr12-quasar-catalog/ Mass data taken from https://arxiv.org/pdf/1609.09489.pdf, data headers are shown below
| Variable | Info |
|---|---|
| RAdeg | Right Ascension in decimal degrees |
| DEdeg | Declination in decimal degrees |
| z | Redshift |
| Mi | Absolute Magnitude i-band |
| L5100 | log10(L at 5100A in erg s^-1) |
| eL5100 | error in log10(L at 5100A in erg s^-1) |
| L3000 | log10(L at 3000A in erg s^-1) |
| eL3000 | error in log10(L at 3000A in erg s^-1) |
| L1350 | log10(L at 1350A in erg s^-1) |
| eL1350 | error in log10(L at 1350A in erg s^-1) |
| MBH_MgII | log10(M_BH in units on M_SUN from MgII) |
| MBH_CIV | log10(M_BH in units on M_SUN from CIV) |
| Lbol | log10(bolometric L in erg s^-1) |
| eLbol | error in log10(bolometric L in erg s^-1) |
| nEdd | log10(Eddington ratio) |
locator.py is modified from https://github.com/kristalynne/tess_agn
Below is some python-ish pseudocode that shows how we generate lightcurves
for row in target_list:
try:
star = eleanor.multi_sectors(sectors='all',tic=0,coords=(ra,dec),tc=True)
full_lc = lightkurve.LightCurve([],[])
for observation in star:
data = eleanor.TargetData(observation,bkg_size=15,do_pca=False)
data.custom_aperture(shape='rectangle', h=1,w=1, method='exact')
data.get_lightcurve()
q = data.quality == 0
lc = data.to_lightkurve()
lc.flux = lc.flux - numpy.median(lc.flux)
full_lc = full_lc.append(lc)
save_lightcurve(sector_lightcurve_path,sector_lightcurve)
save_lightcurve(full_lightcurve_path,full_lc)
except:
terrible_targets.append(row)
continuetest