last updated December 2012.
This briefly describes the light curve data used for my PhD thesis at JHU. I provide all the light curves extracted from these data as an SQLite database, a few CSV files summarizing the properties of the variable stars we found, and a couple of Python modules to help get light curves out of the database.
The full database is 14 GB gzipped. Get it here: stripe82-combined-db-with-probvars.sqlite.gz. You'll need the sqlite3 command-line binary to read this (after un-gzipping it). Alternatively, use Python, which comes with sqlite3 support built-in.
See the thesis PDF for details about how this catalog was used to find and characterize M-dwarf eclipsing binaries from the sparsely sampled SDSS Stripe 82 time-series photometry. A host of other variable objects were also detected, and the data are probably useful for other variability studies as well. See the sections below for hints on how to use the data.
How to get a light curve for an object
I've provided a couple of Python modules that will do this for you, given that you have the un-gzipped SQLite database on hand, along with Numpy installed in your Python environment:
Use stripe82_lcutils.py to get the light curve for a particular object, e.g.:
>>> import stripe82_lcutils as s82
>>> lcdict = s82.get_db_lc('34146-0302-0113',
'stripe82-combined-db-with-probvars.sqlite',
lc_type=1,
sigclip=4.0)
This returns the normalized ensemble differential magnitude light-curve for the object 34146-0302-0113. Use lc_type = 2 to get the de-reddened SDSS PSF mag light curve and lc_type = 3 to get the de-reddened SDSS fiber mag light curve. A dictionary is returned with MJDs and magnitude measurements for each SDSS band.
>>> mjd_r, norm_mag_r, err_r = (lcdict['mjd_r'],
lcdict['norm_mag_r'],
lcdict['err_r'])
This gives you the MJD, mag, and mag measurement error timeseries in the SDSS r band for this particular object.
The periodic variable catalog and light curves
We found 1,162 periodic variable stars in these data. A list of these is in the file: stripe82-periodic-variables-jun-2012.csv. The columns in this file are:
uid TEXT PRIMARY KEY, <-- unique ID of the object vartype TEXT, <-- variable type (EB, ELL, RRab, RRc, HADS, rotation, unknown) period DOUBLE <-- period in days
The epoch used to fold each periodic variable light curve was the Modified Julian Date (MJD) of the faintest light curve point in the SDSS r band; this isn't in the table, but can be obtained in a fairly straight-forward way. A better way to determine this would be to do a spline fit to the phased light curve, find the minimum, get the MJD of the fit closest to that, and use that as the epoch of the periodic variability.
Light curves for the 1,162 periodic variables we found in these data are available as a gzipped tarball: stripe82-periodicvar-lcs.tar.gz.
These are all CSVs with names
like: {UID}-{vartype}-{period}.csv, for
example: 26412-0300+0050-HADS-0.04066.csv. Each file has
the following format:
mjdu,mjdg,mjdr,mjdi,mjdz,
norm_diffmag_u,norm_diffmag_g,norm_diffmag_r,norm_diffmag_i,norm_diffmag_z,
diffmag_err_u,diffmag_err_g,diffmag_err_r,diffmag_err_i,diffmag_err_z,
diffsdss_u,diffsdss_g,diffsdss_r,diffsdss_i,diffsdss_z
where:
mjd_Xis the modified Julian Date (Julian - 2400000.5) for each SDSS band: u, g, r, i, znorm_diffmag_Xis the differential magnitude produced by our pipeline, normalized to the median (so centered about zero) for each SDSS band: u, g, r, i, zdiffmag_err_Xis the error for each normalized differential magnitude measurement in each SDSS band: u, g, r, i, zdiffsdss_Xis the normalized differential magnitude added back to the median SDSS PSF magnitude for each SDSS band: u, g, r, i, z. This is mostly if you want a light curve that reflects the actual magnitudes of the object in each band instead of a light curve that's centered about zero.
The variable point source catalog
We found 45,822 probable variable objects in these data. A gzipped CSV of these objects can be found at (15 MB): stripe82-probable-variables-jun-2012.csv.gz. Ignore the first line when reading it in; it's the header. The columns for the CSV are listed below:
uid TEXT, <-- unique identifier of the object sdss_name TEXT, <-- SDSS name of the object ra REAL, <-- right ascension in decimal degrees dec REAL, <-- declination in decimal degrees nobs INT, <-- number of total detections of the object object_tags TEXT, <-- object type tags classified by SEGUE color-cuts med_dered_u REAL, <-- median dereddened u-band magnitude of the object med_dered_g REAL, <-- median dereddened g-band magnitude of the object med_dered_r REAL, <-- median dereddened r-band magnitude of the object med_dered_i REAL, <-- median dereddened i-band magnitude of the object med_dered_z REAL, <-- median dereddened z-band magnitude of the object med_ug REAL, <-- median u-g color of the object med_gr REAL, <-- median g-r color of the object med_ri REAL, <-- median r-i color of the object med_iz REAL, <-- median i-z color of the object stdev_dered_u REAL, <-- RMS of dereddened u-band magnitude of the object stdev_dered_g REAL, <-- RMS of dereddened g-band magnitude of the object stdev_dered_r REAL, <-- RMS of dereddened r-band magnitude of the object stdev_dered_i REAL, <-- RMS of dereddened i-band magnitude of the object stdev_dered_z REAL, <-- RMS of dereddened z-band magnitude of the object stet_ug REAL, <-- Stetson variability index using u and g bands stet_gr REAL, <-- Stetson variability index using g and r bands stet_ri REAL, <-- Stetson variability index using r and i bands stet_iz REAL, <-- Stetson variability index using i and z bands varflag_u NUM, <-- set if object > 4-sigma above u RMS-mag curve varflag_g NUM, <-- set if object > 4-sigma above g RMS-mag curve varflag_r NUM, <-- set if object > 4-sigma above r RMS-mag curve varflag_i NUM, <-- set if object > 4-sigma above i RMS-mag curve varflag_z NUM, <-- set if object > 4-sigma above z RMS-mag curve skew_u REAL, <-- skew of the ensemble diff mag u-band light curve skew_g REAL, <-- skew of the ensemble diff mag g-band light curve skew_r REAL, <-- skew of the ensemble diff mag r-band light curve skew_i REAL, <-- skew of the ensemble diff mag i-band light curve skew_z REAL, <-- skew of the ensemble diff mag z-band light curve kurt_u REAL, <-- kurtosis of the diff mag u-band light curve kurt_g REAL, <-- kurtosis of the diff mag g-band light curve kurt_r REAL, <-- kurtosis of the diff mag r-band light curve kurt_i REAL, <-- kurtosis of the diff mag i-band light curve kurt_z REAL, <-- kurtosis of the diff mag z-band light curve sig_ens_mag_u REAL, <-- RMS of ensemble diff u-band light curve sig_ens_mag_g REAL, <-- RMS of ensemble diff u-band light curve sig_ens_mag_r REAL, <-- RMS of ensemble diff u-band light curve sig_ens_mag_i REAL, <-- RMS of ensemble diff u-band light curve sig_ens_mag_z REAL <-- RMS of ensemble diff u-band light curve
The full point source catalog
There are 2,000,241 objects in total. A summary of these objects' SDSS properties can be found in object summary CSV file (582 MB): stripe82-all-point-sources-jun-2012.csv.gz. This is a gzipped CSV file. The first line is the header, which should be skipped when you read the file in. The CSV has the following columns:
uid TEXT PRIMARY KEY, <-- this is the unique ID for each object sdss_name TEXT, <-- this is the SDSS JXXXXXX.XX+-XXXXXX.X object name ra DOUBLE, <-- right ascension in decimal degrees dec DOUBLE, <-- declination in decimal degrees nobs INTEGER, object_tags TEXT, <-- type of object classified by SDSS/SEGUE color-cuts med_orig_u DOUBLE, <-- median SDSS u-band light curve magnitude med_orig_g DOUBLE, <-- median SDSS g-band light curve magnitude med_orig_r DOUBLE, <-- median SDSS r-band light curve magnitude med_orig_i DOUBLE, <-- median SDSS i-band light curve magnitude med_orig_z DOUBLE, <-- median SDSS z-band light curve magnitude med_dered_u DOUBLE, <-- median SDSS de-reddened u-band light curve magnitude med_dered_g DOUBLE, <-- median SDSS de-reddened g-band light curve magnitude med_dered_r DOUBLE, <-- median SDSS de-reddened r-band light curve magnitude med_dered_i DOUBLE, <-- median SDSS de-reddened i-band light curve magnitude med_dered_z DOUBLE, <-- median SDSS de-reddened z-band light curve magnitude err_med_u DOUBLE, <-- standard error of the median u-band magnitude err_med_g DOUBLE, <-- standard error of the median g-band magnitude err_med_r DOUBLE, <-- standard error of the median r-band magnitude err_med_i DOUBLE, <-- standard error of the median i-band magnitude err_med_z DOUBLE, <-- standard error of the median z-band magnitude stdev_dered_u DOUBLE, <-- RMS of the dereddened u-band light curve stdev_dered_g DOUBLE, <-- RMS of the dereddened g-band light curve stdev_dered_r DOUBLE, <-- RMS of the dereddened r-band light curve stdev_dered_i DOUBLE, <-- RMS of the dereddened i-band light curve stdev_dered_z DOUBLE, <-- RMS of the dereddened z-band light curve med_ug DOUBLE, <-- median u-g color of the object med_gr DOUBLE, <-- median g-r color of the object med_ri DOUBLE, <-- median r-i color of the object med_iz DOUBLE, <-- median i-z color of the object err_med_ug DOUBLE, <-- standard error of the u-g color err_med_gr DOUBLE, <-- standard error of the u-g color err_med_ri DOUBLE, <-- standard error of the u-g color err_med_iz DOUBLE, <-- standard error of the u-g color stdev_dered_ug DOUBLE, <-- RMS of the dereddened u-g color stdev_dered_gr DOUBLE, <-- RMS of the dereddened g-r color stdev_dered_ri DOUBLE, <-- RMS of the dereddened r-i color stdev_dered_iz DOUBLE, <-- RMS of the dereddened i-z color stet_ug DOUBLE, <-- Stetson variability index using u and g bands stet_gr DOUBLE, <-- Stetson variability index using g and r bands stet_ri DOUBLE, <-- Stetson variability index using r and i bands stet_iz DOUBLE, <-- Stetson variability index using i and z bands time_sparsity DOUBLE <-- length of light curve in JD / number of observations
