Scribd
Upload
31 views4 pages

Polynomials in MATLAB: P (X) X 3x

The document discusses using polynomials in MATLAB, including evaluating polynomials, finding roots of polynomials, polynomial interpolation, and parametric curves. Key topics covered include using polyval to evaluate polynomials, roots to find roots, poly to calculate coefficients from roots, and polyfit for interpolation.

Uploaded by

معتز انور
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
31 views4 pages

Polynomials in MATLAB: P (X) X 3x

The document discusses using polynomials in MATLAB, including evaluating polynomials, finding roots of polynomials, polynomial interpolation, and parametric curves. Key topics covered include using polyval to evaluate polynomials, roots to find roots, poly to calculate coefficients from roots, and polyfit for interpolation.

Uploaded by

معتز انور
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Math 436

Fall 2008
November 5
Polynomials in MATLAB

1 Polynomial Evaluation
A polynomial is completely known by its coefficients! For example the polynomial
p(x) = x3 − 3x2 + 1
has coefficients, beginning with the largest power, a1 = 1, a2 = −3, a3 = 0 and a4 = 1. Once these
numbers are given, it is plain that p(1) = −1 or p(4) = 17 or p(−2) = −19. MATLAB lets you
compute polynomials with the polyval command. It has the structure
polyval(coefficient vector, input data)
In our case, we could enter
>> a = [1 -3 0 1]
a =
1 -3 0 1
>> polyval(a,[1 4 -2])
ans =
-1 17 -19
As you can see, we can calculate the three values simultaneously entering a vector as the second
argument. For a plot of the curve on the interval [−1, 4] one could proceed as follows:
>> a = [1 -3 0 1]
a =
1 -3 0 1
>> x=linspace(-1,4); % generate x data
>> p = polyval(a,x); % calculate polynomial values
>> plot(x,p); grid on % graph the function
20

15

10

−5
−1 −0.5 0 0.5 1 1.5 2 2.5 3 3.5 4
2 Roots
It is known that a polynomial of degree n has exactly n, possibly complex roots. Here you have to
account for the multiplicity of the roots as well. The roots command determines the roots from
the coefficients. For our polynomial p(x) = x3 − 3x2 + 1 we find

>> roots([1 -3 0 1])


ans =
2.8794
0.6527
-0.5321

which is consistent with the information from the graph.


The reverse of the roots command is poly. Here you can easily calculate the coefficients of
a polynomial with roots at desired locations. Suppose you want to find a polynomial with simple
roots at −1 and 4, and a double root at 2. With poly you just need to enter

>> poly([-1 4 2 2])


ans =
1 -7 12 4 -16

The desired polynomial is p(x) = x4 − 7x3 + 12x2 + 4x − 16. A graph confirms the result:

>> a = poly([-1 4 2 2]);


>> x = linspace(-2,5);
>> p = polyval(a,x);
>> plot(x,p), grid on
100

80

60

40

20

−20
−2 −1 0 1 2 3 4 5

3 Polynomial Interpolation
Polynomial interpolation is built into MATLAB. The command is

a = polyfit(x data, y data, degree of the polynomial)

It returns the coefficients of the interpolating polynomial. From class there is exactly one polynomial
of degree ≤ N − 1 which passes through N given points. MATLAB lets you approximate with lower
degree polynomials in the least squares sense. For instance if you take the degree to be one, it will
find the least squares regression line. Our usual example with the data

x 0 1 2 4
y 1 2 4 16

can be treated as follows (I switched to format rat)

>> X = [0 1 2 4], Y = [1 2 4 16]


X =
0 1 2 4
Y =
1 2 4 16
>> a = polyfit(X,Y,3)
a =
5/24 -1/8 11/12 1
>> polyval(a,3) % p(3)
ans =
33/4

The vector a contains the coefficients of the polynomial, i.e. the interpolating polynomial is P4 (x) =
5 2 1 2 11
24 x − 8 x + 12 x + 1 and its value at 3 is P4 (3) = 33/4 = 8.25. For a graph of the interpolating
polynomial along with the underlying function f (x) = 2x and the data points on the interval [−1, 5]
continue as follows

>> x=linspace(-1,5); % generate x values


>> f = 2.^x; % underlying function
>> p = polyval(a,x); % interpolating polynomial
>> plot(X,Y,’o’,x,f,x,p), grid on % graph

35

30

25

20

15

10

−5
−1 0 1 2 3 4 5

3.1 Least Squares Line


Suppose you want to find the least squares regression line for the data. Here you are looking to a
polynomial of degree one to approximate the data X and Y, and we enter 1 for the degree in polyfit
(here it is assumed the definitions for the vectors X and Y are still in effect).
>> b = polyfit(X,Y,1)
b =
3.8571 -1.0000
>> y = polyval(b,x); % for a plot
>> plot(x,y,X,Y,’o’), grid on

Thus the least square regression line is y = 3.8571x − 1.


20

15

10

−5
−1 0 1 2 3 4 5

4 Parametric Curves
For parametric curves (x(t), y(t) you need to display the x-values versus the y with t as a parameter.
For example, the curve x(t) = t3 and y(t) = t2 with −1 ≤ t ≤ 1 can be generated as follows:

>> t = linspace(-1,1);
>> x = polyval([1 0 0 0],t);
>> y = polyval([1 0 0],t);
>> plot(x,y), grid on, axis(’equal’)

1.2

0.8

0.6

0.4

0.2

−0.2

−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1

√
3
This, of course, is the graph of the function y = x2 .

You might also like