Scribd Logo
Upload

Welcome to Scribd!

  • 31 views

    Computational Methods in Physics: Seminar

    Uploaded by

    Krishna Agarwal
    This seminar discusses Muller's method, an iterative method for finding roots of equations. [1] Muller's method approximates the function as a quadratic equation near the root. [2] It converges quadratically, faster than secant or bisection methods. [3] The method finds roots without needing to compute derivatives, overcoming a limitation of Newton-Raphson's method.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Computational Methods in Physics: Seminar

    Uploaded by

    Krishna Agarwal
    31 views4 pages
    This seminar discusses Muller's method, an iterative method for finding roots of equations. [1] Muller's method approximates the function as a quadratic equation near the root. [2] It converges quadratically, faster than secant or bisection methods. [3] The method finds roots without needing to compute derivatives, overcoming a limitation of Newton-Raphson's method.

    Original Title

    seminar NEW.docx

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This seminar discusses Muller's method, an iterative method for finding roots of equations. [1] Muller's method approximates the function as a quadratic equation near the root. [2] It converges quadratically, faster than secant or bisection methods. [3] The method finds roots without needing to compute derivatives, overcoming a limitation of Newton-Raphson's method.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    31 views4 pages

    Computational Methods in Physics: Seminar

    Uploaded by

    Krishna Agarwal
    This seminar discusses Muller's method, an iterative method for finding roots of equations. [1] Muller's method approximates the function as a quadratic equation near the root. [2] It converges quadratically, faster than secant or bisection methods. [3] The method finds roots without needing to compute derivatives, overcoming a limitation of Newton-Raphson's method.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 4

    SEMINAR

    ABSTRACT
    COMPUTATIONAL METHODS IN PHYSICS
    18MSPH2H04
    Krishna Agarwal
    MULLER’S METHOD

    In this method f(x) is approximated by a second degree curve in the vicinity of root. The roots of
    quadratic equation are then assumed to be approximation of roots f(x) = 0. This method is iterative

    And converges quadratically, and can be used to obtain complex roots.

    This is an iteration method just like secant method unlike secant there are 3 distinct approximations.

    The rate of convergence, i.e., how much closer we move to the root at each step, is approximately 1.84
    in Muller Method, whereas it is 1.62 for secant method, and linear, i.e., 1
    for both Regula – falsi Method and bisection method . So, Muller Method
    is faster than Bisection, Regula – Falsi and Secant method. Although, it is
    slower than Newton – Raphson’s Method, which has a rate of
    convergence of 2, but it overcomes one of the biggest drawbacks of
    Newton-Raphson Method, i.e., computation of derivative at each step.

    Let x , x i−1 , x i be 3 distinct approximations of a second degree curve


    i−2

    f ( x )=0

    Let y i−2 , y i−1 , yi be corresponding values of y=f (x ). Now any second degree curve passing trough
    point ( x i , y i ) can be written as

    p ( x ) =A (x−x i )2 + B ( x−x i ) + y i---------------- (1)

    Since the curve also pass through ( x i−1 , y i−1) and( x i−2 , y i−2) , we get

    y i−2= A( x−xi −2 )2+ B ( x−x i−2 ) + y i----------- (2)

    y i−1= A(x −xi−1 )2+ B ( x−x i−1 ) + y i ---------- (3)

    On solving equation (2) and (3) we can find value of A and B

    y i−1− y i y i−2− y i
    A= +
    (x i−1 −xi−2 )(x i−1 −xi ) (x i−2−x i )( xi−2−x i−1 )

    And,

    y i−1− y i
    B= − A(x i−1−x i)
    x i−1−x i
    Substituting these two values in equation (1) we’ll get our 1 st approximation x i+1

    2 yi
    x i+1=x i−
    B± √ ( B ¿ ¿ 2 ¿−4 A y i) ¿ ¿

    The sign in denominator should be chosen so the denominator will be largest in magnitude. With this
    choice the next approximation root will have low error.

    C-Program
    #include<stdio.h>
    #include<math.h>
    #define I 2
    float f(float x)
    {
    return x*x*x-2*x-5;
    }
    main ()
    {
    int i, itr, maxmitr;
    float x[4], li, di, mu, s, l, allerr;
    printf("\nEnter the three initial guesses:\n");
    for (i=I-2; i<3; i++)
    scanf("%f", &x[i]);
    printf("Enter allowed error and maximum iterations:\n");
    scanf("%f %d", &allerr, &maxmitr);
    for (itr=1; itr<=maxmitr; itr++)
    {
    li = (x[I] - x[I-1]) / (x[I-1] - x[I-2]);
    di = (x[I] - x[I-2]) / (x[I-1] - x[I-2]);
    mu = f(x[I-2])*li*li - f(x[I-1])*di*di + f(x[I])*(di+li);
    s = sqrt ((mu*mu - 4*f(x[I])*di*li*(f(x[I-2])*li - f(x[I-1])*di + f(x[I]))));
    if (mu < 0)
    l = (2*f(x[I])*di)/(-mu+s);
    else
    l = (2*f(x[I])*di)/(-mu-s);
    x[I+1] = x[I] + l*(x[I] - x[I-1]);
    printf("At iteration no. %3d, x = %7.5f\n", itr, x[I+1]);
    if (fabs (x[I+1] - x[I]) < allerr)
    {
    printf("After %3d iterations, the required root is %6.4f\n", itr, x[I+1]);
    return 0;
    }
    for (i=I-2; i<3; i++)
    x[i] = x[i+1];
    }
    printf("The required solution does not converge or iterations are insufficient\n");
    return 1;
    }
    References

     Introductory methods of numerical analysis by- S.S sastry


     www.tutorialspoint.com

    You might also like