Plotting and helper functions

Plotting

This module contains functions to plot light curves and detrending results.

democratic_detrender.plot.plot_detrended_lc(xs, ys, detrend_labels, t0s_in_data, window, period, colors, duration, mask_width=1.1, depth=None, figname=None, title=None)

inputs: x = times ys = [detrended light curves] of length N number of detrendings detrend_labels = [detrending type] of length N number of detrendings t0s_in_data = midtransits in data window = what fraction of the period to plot on either side of transit (ie. window=1/2 means 1/2 period on either side) period = planet period to define plotting limit colors = [colors] of length N number of detrendings figname = Name of file if you want to save figure

return: None

democratic_detrender.plot.plot_individual_outliers(time, flux, time_out, flux_out, t0s, period, window, depth, figname)

input:

time = array of time values flux = array of flux values time_out = array of time values without outliers flux_out = array of flux values without outliers period = planet period to define plotting limit window = what fraction of the period to plot on either side of transit (ie. window=1/2 means 1/2 period on either side) t0s = midtransits in data

democratic_detrender.plot.plot_outliers(time, flux, time_out, flux_out, moving_median, kepler_quarters, figname, object_id)

input:

time = array of time values flux = array of flux values time_out = array of time values without outliers flux_out = array of flux values without outliers

democratic_detrender.plot.plot_phase_fold_lc(time, lc, period, t0s, xlim, figname)

democratic_detrender.plot.plot_split_data(x_split, y_split, t0s, figname, object_id)

democratic_detrender.plot.plot_transit(xs_star, ys_star, xs_transit, ys_transit, t0, period, title, bin_window, object_id, problem_times_input=None, dont_bin=False)

democratic_detrender.plot.plot_transits(x_transits, y_transits, mask_transits, t0s, period, bin_window, object_id, problem_times_input=None, dont_bin=False, data_name=None)

Helper functions

This module contains helper functions for statistical calculations and data processing.

democratic_detrender.helper_functions.bin_data(xs, ys, window)

democratic_detrender.helper_functions.determine_cadence(times)

democratic_detrender.helper_functions.durbin_watson(residuals)

Calculate the Durbin-Watson statistic for a given set of residuals.

Parameters:

residuals – array-like, residuals from a model

Returns:

Durbin-Watson statistic

Return type:

float

democratic_detrender.helper_functions.ensemble_step(y_detrended, yerr_detrended=None, method='median')

Combine multiple detrending methods into a single flux time series (method-marginalized light curve) and propagate uncertainties.

Parameters:

• y_detrended (array-like, shape (n_times, n_methods)) – Detrended flux values. Each column is the SAME target light curve after a different detrending method. Rows are timestamps. Can include NaNs.

• yerr_detrended (array-like, shape (n_times,)) – The per-point uncertainty for the target light curve BEFORE method-marginalization. If None, it’s treated as zeros. (This matches how your snippet uses yerr_detrended as a 1D array.)

• method ({"median", "mean"}) –

  How to combine across methods at each timestamp: - “median”: use np.nanmedian across methods,

  and use MAD as the method-scatter term.

  • ”mean”: use np.nanmean across methods,

    and use standard deviation as the method-scatter term.

Returns:

• flux_mm (ndarray, shape (n_times,)) – Method-marginalized flux (the final combined light curve).

• flux_err_mm (ndarray, shape (n_times,)) – Updated per-point uncertainty after marginalizing over method. This is:

  sqrt( yerr_detrended^2 + scatter_term^2 )

  where scatter_term is MAD (for median) or std (for mean), both computed across methods at that timestamp.

Notes

• We transpose internally only to mirror your original code, where you did y_detrended.T then axis=1. After transpose we have shape (n_methods, n_times), so axis=0 would also work, but we’ll stick to axis=0/axis=1 choices that reproduce your math exactly.

• MAD uses scale=1/1.4826 to match your exact call, i.e. the raw unscaled MAD, not the Gaussian-scaled MAD.

democratic_detrender.helper_functions.find_nearest(array, value)

democratic_detrender.helper_functions.get_detrended_lc(y, detrending_model)

Get detrended light curve (LC).

Parameters:

• y (array) – Light curve (LC).

• detrending_model (array) – Stellar detrending model evaluated at the same time as LC.

Returns:

Detrended light curve evaluated at the same time as input LC.

Return type:

array