Math-MatrixLUP/examples/polynomial_iterpolation.pl

#!/usr/bin/perl

# Polynomial interpolation:
#   find the polynomial of lowest possible degree that passes through all the points of a given dataset.

# See also:
#   https://en.wikipedia.org/wiki/Vandermonde_matrix
#   https://en.wikipedia.org/wiki/Polynomial_interpolation

use 5.014;
use lib qw(../lib);
use experimental qw(signatures);

use Math::MatrixLUP;
use Math::AnyNum qw(ipow sum);

# A sequence of n numbers
my @v = (35, 85, 102, 137, 120);

# Create a new nXn Vandermonde matrix
my $A = Math::MatrixLUP->build(scalar(@v), sub ($i, $j) {
        ipow($i, $j);
    }
);

my $S = $A->solve(\@v);

say "Coefficients: [", join(', ', @$S), "]";
say "Polynomial  : ", join(' + ', map { "($S->[$_] * x^$_)" } 0 .. $#{$S});
say "Terms       : ", join(', ', map {my $x = $_; sum(map { $x**$_ * $S->[$_] } 0 .. $#{$S})} 0 .. $#v);

__END__
Coefficients: [35, 455/4, -2339/24, 155/4, -121/24]
Polynomial  : (35 * x^0) + (455/4 * x^1) + (-2339/24 * x^2) + (155/4 * x^3) + (-121/24 * x^4)
Terms       : 35, 85, 102, 137, 120