compute_modshift_metrics
- exovetter.modshift.compute_modshift_metrics(time, flux, model, period_days, epoch_days, duration_hrs, show_plot=True)[source]
Compute Jeff Coughlin’s Modshift metrics.
This algorithm is adapted from the Modshift code used in the Kepler Robovetter and documented on https://exoplanetarchive.ipac.caltech.edu/docs/KSCI-19105-002.pdf
(see page 30, and appropriate context)
Jeff uses modshift to refer to both the transit significance tests, as well as a suite of other, related, tests. This code only measures the metrics for the transit significance measurements.
The algorithm is as follows:
Fold and bin the data
Convolve binned data with model
Identify the three strongest dips, and the strongest positive excursion
Remove some of these events, and measure scatter of the rest of the data
Scale the convolved data by the per-point scatter so that each point represents the statistical significance of a transit signal at that phase.
Record the statistical significance of the 4 events.
- Parameters:
- time
(1d numpy array) times of observations in units of days
- flux
(1d numpy array) flux values at each time. Flux should be in fractional amplitude (with typical out-of-transit values close to zero)
- model
(1d numpy array) Model transit flux based on the properties of the TCE len(model) == len(time)
- period_days, epoch_days, duration_hrsfloat
Properties of the transit
- show_plotbool
Display plot. This needs
matplotlib
to be installed.
- Returns:
- resultsdict
- Raises:
- ValueError
Invalid inputs.