Today’s class • Roots of Equations • Polynomials Numerical Methods, Lecture 6 1 Prof. Jinbo Bi CSE, UConn
Mar 15, 2016
Today’s class• Roots of Equations• Polynomials
Numerical Methods, Lecture 6 1 Prof. Jinbo Bi
CSE, UConn
• Locate the roots of a set of simultaneous equations
• Use extensions of the various open methods
• Fixed-point iteration is not practical for nonlinear systems
Numerical Methods, Lecture 6 2 Prof. Jinbo Bi
CSE, UConn
Systems of nonlinear equations
• Using Newton-Raphson method• First order Taylor approximation
• Solve for xi+1 and yi+1
Systems of nonlinear equations
yf
yyxf
xxyxfyxf iii
iiiii
,1
1,1
1,11,1 )()(),(),(
yf
yyxf
xxyxfyxf iii
iiiii
,2
1,2
1,21,2 )()(),(),(
Numerical Methods, Lecture 6 3 Prof. Jinbo Bi
CSE, UConn
Numerical Methods, Lecture 6 4 Prof. Jinbo Bi
CSE, UConn
Systems of nonlinear equations
• Newton-Raphson method can still diverge as with the single equation method
• Can be generalized to more than two equations - requires solving larger linear systems
Numerical Methods, Lecture 6 5 Prof. Jinbo Bi
CSE, UConn
Systems of nonlinear equations
Polynomials• A polynomial is of the form:
• The roots of a polynomial follow these rules:• There will be n roots to an n-th order
polynomial. The roots may be real or complex and need not be distinct
• If n is odd, there is at least one real root• If complex roots exist, they exist in conjugate
pairs (λ+μi and λ-μi)Numerical Methods, Lecture 6 6 Prof. Jinbo Bi
CSE, UConn
Polynomials• Uses in electrical engineering
• Solving for poles and zeros in transfer functions
• Solving characteristic equations in linear ordinary differential equations
Numerical Methods, Lecture 6 7 Prof. Jinbo Bi
CSE, UConn
Computing with Polynomials
Numerical Methods, Lecture 6 8 Prof. Jinbo Bi
CSE, UConn
Computing with Polynomials
Numerical Methods, Lecture 6 9 Prof. Jinbo Bi
CSE, UConn
Polynomial Deflation• Given a polynomial and a single known
root, deflating the polynomial can reduce the order of the polynomial and remove possible redundant roots
• Example:
• Root x=4 is known
Numerical Methods, Lecture 6 10 Prof. Jinbo Bi
CSE, UConn
Polynomial Deflation
Numerical Methods, Lecture 6 11 Prof. Jinbo Bi
CSE, UConn
• Synthetic Division
Polynomial Deflation
Numerical Methods, Lecture 6 12 Prof. Jinbo Bi
CSE, UConn
Roots of polynomials• Previous methods (Newton-Raphson,
bracketing, etc.) have a few problems when applied to polynomials• Must determine an initial guess of the root• Can not find complex roots• May not converge
• Polynomial-specific methods• Müller’s method• Bairstow‘s method
Numerical Methods, Lecture 6 13 Prof. Jinbo Bi
CSE, UConn
• Similar idea to secant method• Instead of projecting a straight line
through two points to estimate the root, project a parabola through three points to estimate the root
Muller’s method
Numerical Methods, Lecture 6 14 Prof. Jinbo Bi
CSE, UConn
Muller’s method
Numerical Methods, Lecture 6 15 Prof. Jinbo Bi
CSE, UConn
• Find a parabola such that it intersects the function at three points x0, x1, and x2
Muller’s method
Numerical Methods, Lecture 6 16 Prof. Jinbo Bi
CSE, UConn
• Solve for c
• Substitute back into set of equations
• Define new set of variables
Muller’s method
Numerical Methods, Lecture 6 17 Prof. Jinbo Bi
CSE, UConn
• Substitute new variables back into equations
• Solve for a and b
Muller’s method
Numerical Methods, Lecture 6 18 Prof. Jinbo Bi
CSE, UConn
• Now that we have found the coefficients of the parabola, find where it intersects the x-axis to get the next root estimate
• The intersection point with the x-axis is simply the root of the parabola which we can find using the quadratic formula
Muller’s method
Numerical Methods, Lecture 6 19 Prof. Jinbo Bi
CSE, UConn
• Which of the two quadratic roots do you choose?• Choose the sign that agrees with the sign of
b, making the denominator the largest• That will bring xr closer to x2
Muller’s method
Numerical Methods, Lecture 6 20 Prof. Jinbo Bi
CSE, UConn
• How do you pick the three points for the next iteration?• If there are only real roots, choose xr and the
closer two of the orignal three roots to xr.• If there are complex roots, use a sequential
approach: use xr, x2, and x1 to replace x2, x1, and x0
Muller’s method
Numerical Methods, Lecture 6 21 Prof. Jinbo Bi
CSE, UConn
• Find roots of f(x)=x3-13x-12• Use x0=4.5, x1=5.5, and x2=5 as the
initial guesses
Example
Numerical Methods, Lecture 6 22 Prof. Jinbo Bi
CSE, UConn
Example
Numerical Methods, Lecture 6 23 Prof. Jinbo Bi
CSE, UConn
Example
Numerical Methods, Lecture 6 24 Prof. Jinbo Bi
CSE, UConn
• Based on dividing the polynomial by (x-t) where t is the root estimate
• If there is no remainder, we found the root
• If not, we adjust the guess
Bairstow’s method
Numerical Methods, Lecture 6 25 Prof. Jinbo Bi
CSE, UConn
• Factor by x-t
• Factor by x2-rx-s
Bairstow’s method
Numerical Methods, Lecture 6 26 Prof. Jinbo Bi
CSE, UConn
• We have to find values of r and s such that the remainder is 0
• The remainder is 0 if b1=b0=0
• b1 and b0 are both functions of r and s, so we can give a Taylor series expansion of both
Bairstow’s method
Numerical Methods, Lecture 6 27 Prof. Jinbo Bi
CSE, UConn
Bairstow’s method
Numerical Methods, Lecture 6 28 Prof. Jinbo Bi
CSE, UConn
• Solve for Δr and Δs• Use the results to adjust r and s and iterate• Stop when the approximate error in r and s is
sufficient
Bairstow’s method
Numerical Methods, Lecture 6 29 Prof. Jinbo Bi
CSE, UConn
• Find roots of f(x)= x5 - 3.5x4 + 2.75x3 + 2.125x5 - 3.875x + 1.25
• Use s= -1, r= -1 as the initial guesses
Example
Numerical Methods, Lecture 6 30 Prof. Jinbo Bi
CSE, UConn
• Solve linear system
• New s and r
Example
Numerical Methods, Lecture 6 31 Prof. Jinbo Bi
CSE, UConn
• Use s= 0.1381, r= -0.6442
Example
Numerical Methods, Lecture 6 32 Prof. Jinbo Bi
CSE, UConn
• Solve linear system
• New s and r
Example
Numerical Methods, Lecture 6 33 Prof. Jinbo Bi
CSE, UConn
• Keep going until s= 0.5 and r= -0.5• Then solve the following:
Example
Numerical Methods, Lecture 6 34 Prof. Jinbo Bi
CSE, UConn
• The quotient coefficients are as follows
• Now we have to solve:
Example
Numerical Methods, Lecture 6 35 Prof. Jinbo Bi
CSE, UConn
• If no s or r are known, start with s=0, r=0• Converges fast• May diverge
Bairstow’s method
Numerical Methods, Lecture 6 36 Prof. Jinbo Bi
CSE, UConn
• Bisection• 2 initial guesses, slow convergence, will not
diverge• False-position
• 2 initial guesses, slow-medium convergence, will not diverge
Roots of equations
Numerical Methods, Lecture 6 37 Prof. Jinbo Bi
CSE, UConn
• Newton-Raphson• 1 initial guess, fast convergence, may
diverge• Secant
• 2 initial guesses, medium-fast convergence, may diverge
Roots of equations
Numerical Methods, Lecture 6 38 Prof. Jinbo Bi
CSE, UConn
• Muller• 3 initial guesses, medium-fast convergence,
may diverge, only polynomials, complex roots
• Bairstow• 2 initial guesses, medium-fast convergence,
may diverge, only polynomials, complex roots
Roots of equations
Numerical Methods, Lecture 6 39 Prof. Jinbo Bi
CSE, UConn
Next class• Linear algebraic equations• Read Chapter 9• HW3, due 09/24
Numerical Methods, Lecture 6 40 Prof. Jinbo Bi
CSE, UConn