Source
Adapted from my own handwritten and hand-typed notes from Year 1 to Year 3 of my EEE degree
Given data points , polynomial interpolation finds the unique polynomial of degree at most that passes through all of them, so . The Lagrange form writes it directly in terms of the data values.
Formula
with the Lagrange polynomials
Each equals 1 at and 0 at every other node, which is why the sum hits every data point.
Error bound
where ranges between and .
Worked example (linear)
Given and , estimate .
Solution:
- ,
Worked example (quadratic)
Add the point .
Solution:
- , ,
- (true value 2.21920)
- error bound: on , so
Notice that adding one point forced us to rebuild every . The Newton Divided Difference Interpolation form avoids this.