Polynomials

Polynomials are used so commonly in algebra, geometry and math in general that Matlab has special commands to deal with them. The polynomial 2x⁴ + 3x³ − 10x² − 11x + 22 is represented in Matlab by the array [2, 3, -10, -11, 22] (coefficients of the polynomial starting with the highest power and ending with the constant term). If any power is missing from the polynomial its coefficient must appear in the array as a zero. 

Here are some examples of the things that Matlab can do with polynomials. I suggest you experiment with the code…
 

Roots of a Polynomial

% Find  roots of polynomial
p = [1, 2, -13, -14, 24];
r = roots(p) 

% Plot the same polynomial (range -5 to 5) to see its roots
x = -5 : 0.1 : 5;
f = polyval(p,x);
plot(x,f)
grid on
 

Find the polynomial from the roots

If you know that the roots of a polynomial are -4, 3, -2, and 1, then you can find the polynomial (coefficients) this way: 

r = [-4 3 -2 1];
p = poly(r)
 

Multiply Polynomials

The command conv multiplies two polynomial coefficient arrays and returns the coefficient array of their product: 

a = [1 2 1];
b = [2 -2];
c = conv(a,b) 

Look (and try) carefully this result and make sure it’s correct.

Divide Polynomials

Matlab can do it with the command deconv, giving you the quotient and the remainder (as in synthetic division). For example:

% a = 2x^3 + 2x^2 - 2x - 2
% b = 2x - 2
a = [2 2 -2 -2];
b = [2 -2]; 

% now divide b into a finding the quotient and remainder
[q, r] = deconv(a,b)
 

You find quotient q = [1 2 1] (q = x² + 2x + 1), and remainder r = [0 0 0 0] (r = 0), meaning that the division is exact, as expected from the example in the multiplication section…
 

First Derivative

Matlab can take a polynomial array and return the array of its derivative:

a = [2 2 -2 -2] (meaning a = 2x³ + 2x² - 2x - 2)
ap = polyder(a) 

The result is ap = [6 4 -2] (meaning ap = 6x² + 4x - 2) 
 

Fitting Data to a Polynomial

If you have some data in the form of arrays (x, y), Matlab can do a least-squares fit of a polynomial of any order you choose to this data. In this example we will let the data be the cosine function between 0 and pi (in 0.01 steps) and we’ll fit a polynomial of order 4 to it. Then we’ll plot the two functions on the same figure to see how good we’re doing. 

clear; clc; close all 

x = 0 : 0.01 : pi;
% make a cosine function with 2% random error on it
f = cos(x) + 0.02 * rand(1, length(x)); 

% fit to the data
p = polyfit(x, f, 4); 

% evaluate the fit
g = polyval(p,x); 

% plot data and fit together
plot(x, f,'r:', x, g,'b-.')
legend('noisy data', 'fit')
grid on
 

Got it?