6.5.7.1.2. statsmodels.sandbox.nonparametric.smoothers.PolySmoother

class statsmodels.sandbox.nonparametric.smoothers.PolySmoother(order, x=None)[source]

Polynomial smoother up to a given order. Fit based on weighted least squares.

The x values can be specified at instantiation or when called.

This is a 3 liner with OLS or WLS, see test. It’s here as a test smoother for GAM

__init__(order, x=None)[source]

6.5.7.1.2.1. Methods

__init__(order[, x])
df_fit() alias of df_model for backwards compatibility
df_model() Degrees of freedom used in the fit.
df_resid() Residual degrees of freedom from last fit.
fit(y[, x, weights])
gram([d])
predict([x])
smooth(*args, **kwds) alias for fit, for backwards compatibility,