v1.10

Added feature stdev_cut to the penalized spline method. The new parameter takes a floating point value. Outliers larger than stdev_cut are removed in each iteration. The threshold is calculated in standard deviations from the mean (after trend fitting). Older versions of wotan had this parameter hardcoded as PSPLINES_STDEV_CUT = 2 . Now it is user variable. If no value is given, it will fall back on the default.

v1.9

Fixed bug in pspline #57 Value error triggered by gaps smaller than break tolerance

v1.7

Printing status text in pspline switched off in the default. To activate, use verbose=True.

v1.6

Added features to the pspline method as suggested by Rachel Fernandes

v1.5

Fixed a bug in pspline where outliers were only clipped in one direction. Thanks to @GijsMulders

v1.4

Fixed a bug which causes NaNs at the first and last data point in some cases

v1.2

FIXED:

• Bug in cosine which caused rounding errors in numpy.linalg.lstsq

CHANGED

• API of transit_mask to use the keyword time (instead of t)
• API of flatten to use the keyword mask instead of weights, and avoid the requirement of inverting masks (see updated example, it is much simpler now)

ADDED

• lowess detrending now available in combination with transit_mask

v1.1

Added transit_mask to mask known transits during detrending

v1.0.7

Added own lowess implementation instead of statsmodels dependency

• faster (based on numba)
• cleaner code (38 sloc instead of 678 sloc in statsmodels)
• convergence to defined level (default 1e-6) instead of a fixed number of robustifying rounds

v1.0.6

Additional methods:

Fitting a model that is a sum of Gaussian bases:

• ridge Ridge regression (L2 loss, Tikhonov regularization)
• lasso LASSO regression
  (L1 loss, Least Absolute Shrinkage Selector Operator, Tibshirani (1996))
• elasticnet Linear regression model trained with 50% L1 and 50% L2 norm regularization