ECE 1010 ECE Problem Solving I

Polynomial Functions
• Since polynomials occur frequently in mathematics and engi-
neering, MATLAB has a collection of very useful functions
for working with them
• The general form of an Nth degree polynomial function is
N N–1 2
f ( x ) = a0 x + a1 x + … + aN – 2 x + aN – 1 x + aN
where degree denotes the highest power of x present
• The polynomial coefficients are 0, a 1, …, a N
• You may be bothered by the fact that the powers of x and
coefficient subscripts appear to be opposite each other
– This is just the notation adopted by MATLAB for express-
ing polynomials; in fact the sum of the coefficient sub-
script and variable exponent are always a constant of N
• A more compact notation for writing the above is to use sum
notation
N

∑
N–n
f(x) = an x
n=0
0
Note: x = 1
Example: A familiar polynomial function is the quadratic poly-
nomial
2
y ( x ) = a0 x + a1 x + a2

Chapter 3: Polynomial Functions 3–12

ECE 1010 ECE Problem Solving I

Polynomial Evaluation
• Using techniques studied thus far we can use MATLAB used
to evaluate f ( x ) over a range of x values as follows:
% Define the range on x
x = x_min:x_step:x_max;
f = a_0*x.^N + a_1*x.^(N-1) + ...
a_Nm2*x.^2 + a_Nm1*x + a_N;
• Taking the quadratic given above, with defined coefficients,
we can write
» a0 = 3; a1 = 2; a2 = 10; % Define the coefficients
» x = 0:20/100:20; % Define the range on x
» f = a0*x.^2 + a1*x + a2; % Evaluate f(x)
• To polynomial evaluation MATLAB comes with the pre-
defined function polyval
» help polyval
POLYVAL Evaluate polynomial.
Y = POLYVAL(P,X), when P is a vector of length N+1
whose elements are the coefficients of a polynomial, Y
is the value of the polynomial evaluated at X.

Y = P(1)*X^N + P(2)*X^(N-1) + ... + P(N)*X + P(N+1)

If X is a matrix or vector, the polynomial is evalu-

ated at all points in X. See also POLYVALM for evalua-
tion in a matrix sense.
• To evaluate the quadratic given above we would write
» x = 0:20/100:20;
» y = polyval([3, 2, 10],x);

Chapter 3: Polynomial Functions 3–13

ECE 1010 ECE Problem Solving I

Polynomial Operations
• In loose terms we can think of the polynomial coefficient
vector, a, as the polynomial itself
• The sum of two polynomials, say

3 2
g ( x ) = a0 x + a1 x + a2 x + a3
2
h ( x ) = b0 x + b1 x + b2

is the sum of coefficients corresponding to like powers of x

s(x) = g(x ) + h(x)

3 2
= a 0 x + ( a 1 + b 0 )x + ( a 2 + b 1 )x + ( a 3 + b 2 )

• For MATLAB evaluation this amounts to simply adding cor-

responding coefficient vectors a and b by first prepending
the b vector with a leading zero, i.e.,
g = [a0, a1, a2, a3];
h = [0, b0, b1, b2]; % prepending only required
s = g + h; % for poly add and subtract
• If one polynomial is a scalar multiple of another, say
g(x ) = k ⋅ f( x)
the corresponding coefficient vectors are related by k, i.e., if
g = polyval(a,x);
f = polyval(b,x);
% Or given g and the fact that b = k*a
f = k*g;
• Suppose we wish to multiply or divide two polynomials

Chapter 3: Polynomial Functions 3–14

ECE 1010 ECE Problem Solving I

• Recall how tedious it is to multiply two polynomials

2 2
g ( x ) = ( a0 x + a1 x + a2 ) ( b0 x + b1 x + b2 )
2 2 2
= a0 x ( b0 x + b1 x + b2 ) + a1 x ( b0 x + b1 x + b2 )
2
+a 2 ( b 0 x + b 1 x + b 2 )
4 3 2
= a 0 b 0 x + ( a 0 b 1 + a 1 b 0 )x + ( a 0 b 2 + a 1 b 1 + a 2 b 0 )x
+( a 1 b 2 + a 2 b 1 )x + a 2 b 2

• If c is the corresponding coefficient vector of g it follows

that
c = [(a0*b0), (a0*b1+a1*b0), (a0*b2+a1*b1+a2*b0),
(a1*b2+a2*b1), (a2*b2)];
• A MATLAB function that performs this operation automati-
cally is conv(a,b) which stands for convolution (you
learn more about this function in Linear Systems Theory –
ECE 3510)
• In the above example we would type
% To find the coefficients of g(x) = a(x)*b(x) we
% would write
g = conv(a,b)
2 2
Example: g ( x ) = ( x + 2x + 3 ) ( 2x + x + 2 )
» a = [1, 2, 3]; b = [2, 1, 2];
» g = conv(a,b)
g =
2 5 10 7 6

Chapter 3: Polynomial Functions 3–15

ECE 1010 ECE Problem Solving I

• g ( x ) is thus
4 3 2
g ( x ) = 2x + 5x + 10x + 7x + 6
• Polynomial division can be accomplished in like fashion
using the function deconv(a,b) which stands for decon-
volution
Example:
4 3 2
g(x) 2x + 5x + 10x + 7x + 6
h ( x ) = ---------- = ---------------------------------------------------------------
b(x) 2
x + 2x + 2
• Hand calculation requires long division of the polynomial
coefficient vectors, including a remainder vector
2
2x + x + 4
2 4 3 2
x + 2x + 2 2x + 5x + 10x + 7x + 6
4 3
2x + 4x + 4x 2
3 2
x + 6x + 7x + 6
3 2
x + 2x + 2x
2
4x + 5x + 6
2
4x + 8x + 8
Remainder – 3x – 2

• To find the quotient and remainder polynomials in MATLAB

write
» g = [2, 5, 10, 7, 6];
» b = [1, 2, 2];
» [h,r] = deconv(g,b);

Chapter 3: Polynomial Functions 3–16

ECE 1010 ECE Problem Solving I

» h
h =
2 1 4 % --> h(x) = 2x^2+x+4
» r
r =
0 0 0 -3 -2 % --> r(x) = -3x-2

Example: Practice! p. 82, #2 & #5

3 2
f 2 ( x ) = x – 6x + 12x – 8
3 2
f 4 ( x ) = x – 5x + 7x – 3

• For #2 we must evaluate f 2 ( x ) – 2f 4 ( x ) on [0, 4]

» x = 0:4/50:4; % Use 51 points
» a_f2 = [1 -6 12 -8];
» a_f4 = [1 -5 7 -3];
» a_p2 = a_f2 - 2*a_f4;
» f_p2 = polyval(a_p2,x);
» plot(x,f_p2)
» title('Practice! p. 82, #2','fontsize',18)
» ylabel('f2(x) - 2f4(x)','fontsize',16)
» xlabel('x','fontsize',16)

Chapter 3: Polynomial Functions 3–17

ECE 1010 ECE Problem Solving I

Practice! p. 82, #2
4

0
f2(x) - 2f4(x)

-2

-4
3 2
( 1 – 2 )x + ( – 6 + 10 )x + ( 12 – 14 )x + ( – 8 + 6 )
3 2
-6 = – x + 4x – 2x – 2

-8

-10
0 0.5 1 1.5 2 2.5 3 3.5 4
x

• For #5 we must evaluate ( f 4 ( x ) ) ⁄ ( x – 1 ) on [0, 4]

» [q_f5,r_f5] = deconv(a_f4,[1 -1]); % Divide the polys
» q_f5
q_f5 =
1 -4 3 % The quotient is a quadratic
» r_f5
r_f5 =
0 0 0 0 % Note there is no remainder!
» f_p5 = polyval(q_f5,x);
» plot(x,f_p5)
» title('Practice! p. 82, #5','fontsize',18)
» ylabel('f4(x)/(x-1)','fontsize',16)
» xlabel('x','fontsize',16)

Chapter 3: Polynomial Functions 3–18

ECE 1010 ECE Problem Solving I

Practice! p. 82, #5
3

2.5 2 We Plot
x – 4x + 3
3 2 This
2 x–1 x – 5x + 7x – 3
3 2
x –x
2
f4(x)/(x-1)

1.5
– 4x + 7x – 3
2
– 4x + 4x
1
3x – 3
3x – 3
0.5
0
0

-0.5

-1
0 0.5 1 1.5 2 2.5 3 3.5 4
x

Roots of Polynomials
• A problem related to evaluating polynomials is solving for
the roots of a polynomial
• Recall that the roots of y = f ( x ) are those values of x where
f( x) = 0
• For polynomials with real number coefficients, the roots may
be real numbers and/or pairs of complex conjugate numbers
– For a third-order polynomial the roots may be either three
real roots or one real root and a complex pair
– For a second-order polynomial the roots may be either two
real roots or a complex conjugate pair

Chapter 3: Polynomial Functions 3–19

ECE 1010 ECE Problem Solving I

• If we plot f ( x ) we find that the real roots correspond to those

locations where the polynomial crosses the x-axis
2
• Recall that for a quadratic y ( x ) = a 0 x + a 1 x + a 2 the
roots are (quadratic formula)
2
– a 1 ± a 1 – 4a 0 a 2
r 1, r 2 = -------------------------------------------
2a 0
2
Note: The roots are complex if 4a 0 a 2 > a 1
• Hand calculation of polynomial roots becomes impractical as
the order increases, hence we often turn to a numerical solu-
tion
• The MATLAB function roots(a) finds the roots of a poly-
nomial with coefficient vector a
4 3 2
Example: f ( x ) = 2x + 5x + 10x + 7x + 6
» p = [2 5 10 7 6];
» r = roots(p)
r =
-1.0000 + 1.4142i
-1.0000 - 1.4142i
-0.2500 + 0.9682i
-0.2500 - 0.9682i
» polyval(p,r)% Check by evaluating p
% at the root locations
ans = 1.0e-013 *
-0.1243 + 0.0089i
-0.1243 - 0.0089i
-0.0533 + 0.0355i
-0.0533 - 0.0355i % Functional values are small

Chapter 3: Polynomial Functions 3–20

ECE 1010 ECE Problem Solving I

• Given the roots of a polynomial 1, r 2, …, r Nwe know that

N N–1
f ( x ) = a0 x + a2 x + … + aN – 1 x + aN
= ( x – r 1 ) ( x – r 2 )… ( x – r N )

• The MATLAB function poly(r) effectively reconstructs

the polynomial coefficient vector, a, given the vector of roots
by repeated polynomial multiplication
Example: Continue the previous example
» poly(r)
ans =
1.0000 2.5000 5.0000 3.5000 3.0000
• What happened?
– The poly() function must assume that a 0 is normalized
to unity
– In the earlier example a 0 = 2 , so to get the correct
answer we multiply by two following the poly() opera-
tion, e.g.,
» 2*poly(r)
ans =
2.0000 5.0000 10.0000 7.0000 6.0000

Examples: Practice! p. 86, #4

Given
5 4 3 2
g 4 ( x ) = x – 3x – 10x + 27x + 10x – 24

Chapter 3: Polynomial Functions 3–21

ECE 1010 ECE Problem Solving I

Determine the real roots of g 4 ( x ) and then plot the polynomial

over the appropriate interval to verify that the polynomial crosses
the x-axis at the real root locations
» a_g4 = [1 -3 -11 27 10 -24];
» roots(a_g4)' % Use transpose to make output compact
ans =
4.0000 -3.0000 2.0000 1.0000 -1.0000» x =
-4:(5+4)/100:5; % Plot 101 points
» g4 = polyval(a_g4,x);
» plot(x,g4);grid
» title('Practice! p. 86, #4','fontsize',18)
» ylabel('g4(x)','fontsize',16)
» xlabel('x','fontsize',16)

Practice! p. 86, #4
600

400

200

0
g4(x)

-200

-400

-600

-800
-4 -3 -2 -1 0 1 2 3 4 5
x

Chapter 3: Polynomial Functions 3–22