Goodness of a polynomial fit to a time series
Usually kind of a stupid thing to do with a time series, but it’s sometimes
somehow informative for time series with large trends.
- Parameters:
- y : array-like
the time series to analyze.
- k : int, optional
the order of the polynomial to fit to y. Default is 1.
- Returns:
RMS error of the fit.
- Return type:
float