gwinferno.interpolation.BasisSpline

class BasisSpline(n_df, knots=None, interior_knots=None, xrange=(0, 1), k=4, normalize=True)

Bases: object

Methods

bases(xs)	bases form the basis spline design matrix evaluated at xs.
eval(xs, coefs)	eval Evaluate basis spline at xs given coefficients
get_coefficients(xs, ys)	computes the coefficients of the basis components given data 1-D data (xs, ys)
norm(coefs)	norm numerically normalizes the spline
project(bases, coefs)	project given a design matrix (or bases) and coefficients, project the coefficients onto the spline

__call__(xs, coefs)

__call__ Evaluate basis spline at xs given coefficients

Args:

xs (array_like): input values to evaluate the basis spline at coefs (array_like): coefficients for the basis components

Returns:

array_like: The linear combination of the basis components evaluated at xs given the coefficients

__init__(n_df, knots=None, interior_knots=None, xrange=(0, 1), k=4, normalize=True)

Class to construct a basis spline (with the M-Spline basis)

Args:

n_df (int): number of degrees of freedom for the spline knots (array_like, optional): array of knots, if non-uniform knot placing is preferred.

Defaults to None.

interior_knots (array_like, optional): array of interior knots,

if non-uniform knot placing is preferred. Defaults to None.

xrange (tuple, optional): domain of spline. Defaults to (0, 1). k (int, optional): order of the spline +1, i.e. cubic splines->k=4. Defaults to 4 (cubic spline). normalize (bool, optional): flag whether or not to numerically normalize the spline. Defaults to True.

bases(xs)

bases form the basis spline design matrix evaluated at xs. All basis values outside the range of the spline are set to zero.

Args:

xs (array_like): input values to evaluate the basis spline at

Returns:

array_like: the design matrix evaluated at xs. shape (N, *xs.shape)

eval(xs, coefs)

eval Evaluate basis spline at xs given coefficients

Args:

xs (array_like): input values to evaluate the basis spline at coefs (array_like): coefficients for the basis components

Returns:

array_like: The linear combination of the basis components evaluated at xs given the coefficients

get_coefficients(xs, ys)

computes the coefficients of the basis components given data 1-D data (xs, ys)

Args:

xs (array_like): The x values of data ys (array_like): The y values of data

Returns:

the coefficients (array_like), interpolated y-values evaluated at xs (array_like), the design matrix evaluated at xs with shape (N, *xs.shape)

norm(coefs)

norm numerically normalizes the spline

Args:

coefs (array_like): coefficients for the basis components

Returns:

float: the normalization factor given the coefficients

project(bases, coefs)

project given a design matrix (or bases) and coefficients, project the coefficients onto the spline

Args:

bases (array_like): The set of basis components or design matrix to project onto coefs (array_like): coefficients for the basis components

Returns:

array_like: The linear combination of the basis components given the coefficients