latqcdtools.math.spline
=============

`_even_knots(xdata, nknots)`
 
    Return a list of nknots evenly spaced knots. 
    
`_random_knots(xdata, nknots, randomization_factor=1, SEED=None)`
 
    Return a list of nknots randomly spaced knots. 
    
`getSpline(xdata, ydata, num_knots=None, edata=None, order=3, rand=False, fixedKnots=None, getAICc=False, natural=False)`
 
    This is a wrapper that calls SciPy spline interpolation methods, depending on your needs. Generally
    this uses scipy.interpolate.splrep, which uses B-splines. If natural=True and edata=None, it will
    use scipy.interpolate.CubicSpline to solve. If natural=True and edata are provided, it will do a
    smoothing spline that attempts to force no curvature at the endpoints, based on the penultimate points. 

    Args:
        xdata (array-like)
        ydata (array-like)
        num_knots (int):
            The number of knots.
        edata (array-like, optional): 
            Error data. Defaults to None.
        order (int, optional):
            Order of the spline. Defaults to 3.
        rand (bool, optional): 
            Use randomly placed knots? Defaults to False.
        fixedKnots (array-like, optional):
            List of user-specified knots. Defaults to None.
        getAICc (bool, optional): 
            Return corrected Aikake information criterion? Defaults to False.
        natural (bool, optional): 
            Try a natural (no change in slope at the endpoints) cubic spline. Defaults to False. 

    Returns:
        callable spline object
        AICc (optionally)
    
`getSplineErr(xdata, xspline, ydata, ydatae, num_knots=None, order=3, rand=False, fixedKnots=None, natural=False)`
 
    Use getSpline to smooth mean and error bars. Create a spline-smooth band from that. 
    
`TBSpline(xdata, ydata, edata=None, knots=None, order=3, naturalLike=False)`

    A class that prepares a splrep and wraps it with splev.