Light-curve analysis
Analyze your own light curve
Paste or drop your photometry, or load a star from the archive. The page searches it for a period, folds it and classifies it. Nothing is uploaded.
⌘Ctrl + Enter from the text box
or try an example
What files it reads
- Columns of time, magnitude and, optionally, its error, separated by commas, semicolons, tabs or spaces. A header row, or a last
#comment line, names them; without one the order is time, magnitude, error. - Time as JD, HJD, BJD or MJD (MJD is shifted to JD). TESS
btjdand Keplerbkjdcolumns are shifted to BJD when the header names them; any other time is folded as given. - A flux column (
flux,pdcsap_flux, …) with no magnitude column is converted to magnitudes relative to its median. - One band at a time: magnitudes from different filters are never mixed.
Try
Period
—d
searching…
Detections
—
used
Baseline
—d
first to last
Amplitude
—mag
5–95 percentile
d
More fold options
—Click a peak to fold at it. Dotted lines: aliases of the sidereal day, the year and the Moon.
—For eclipses and transits · positive depth = dimming
Searching…
Candidate periods
each tagged with its relation to the fold on screen · click one to fold at itClassify the fold
The archive's classifier, run on the light curve as it is folded above.
downloads a 2 MB model and a 3 MB runtime once, then re-runs whenever the fold changes
Here the model sees one band, with no color, distance or crossmatch, so it is less sure than the archive's own classes. Contact and detached binaries, RR Lyrae and long-period variables are its safest calls.
Variability statistics
Variability statistics
Independent of the fold: brightness, shape, noise and timing, computed exactly as the archive computes them.