Fundamentals of Minterpy#

This section provides a detailed explanation of the mathematical foundation behind Minterpy.

The problem

Multidimensional Polynomial Interpolation

The polynomial interpolation problem is central to Minterpy.

Revisit the polynomial interpolation problem and its connection to approximation on this page.

Multidimensional Polynomial Interpolation

Representing polynomials

Polynomial Bases

Polynomials can be expressed as a linear combination of a set of polynomials called the basis polynomials.

Review the polynomial bases supported by Minterpy on this page.

Multidimensional Polynomial Bases

Getting interpolating polynomials

Interpolation at Unisolvent Nodes

With a careful consideration, a polynomial that interpolate a given function can be constructed.

Review the essential components required to build an interpolating polynomial on this page.

Interpolation at Unisolvent Nodes

Evaluating polynomials

Evaluation of Polynomials

Once a polynomial is obtained, how can it be evaluated at an arbitrary query point?

The answer depends on the basis in which the polynomial is represented.

Evaluation of Multidimensional Polynomials

Changing polynomial basis

Transformation between Bases

A polynomial expressed in one basis may be expressed in another. Depending on the purpose, some representations may be better than the others.

On this page, the theorems behind such a transformation are presented.

Transformation Between Bases

Polynomial based on scattered data

Polynomial Regression

Stable polynomial interpolation requires a careful selection of interpolation points; but what if scattered data is provided instead?

Using the least squares method, you can still construct a polynomial from the data.

Polynomial Regression

Advanced topics

Divided Difference Scheme (DDS)
Multidimensional Divided Difference Scheme (DDS)
Barycentric Transformation
Barycentric Transformation
The Notion of Unisolvence
The Notion of Unisolvence