Scribd Logo
Upload

Welcome to Scribd!

  • 140 views

    Lecture 3 - Errors in Numerical Computation

    Uploaded by

    Huzaika Matloob
    - Numerical methods yield approximate results that contain errors compared to the true analytical solution. - There are different types of errors including absolute error, relative error, and approximation error. - Approximation error includes truncation error from truncating infinite sums, and round-off error from representing numbers with a finite number of bits. - Relative and approximation errors are typically used to analyze errors in numerical solutions since the true value is unknown for real-world problems.

    Copyright:

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

    Lecture 3 - Errors in Numerical Computation

    Uploaded by

    Huzaika Matloob
    140 views15 pages
    - Numerical methods yield approximate results that contain errors compared to the true analytical solution. - There are different types of errors including absolute error, relative error, and approximation error. - Approximation error includes truncation error from truncating infinite sums, and round-off error from representing numbers with a finite number of bits. - Relative and approximation errors are typically used to analyze errors in numerical solutions since the true value is unknown for real-world problems.

    Copyright

    © © All Rights Reserved

    Available Formats

    PPSX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    - Numerical methods yield approximate results that contain errors compared to the true analytical solution. - There are different types of errors including absolute error, relative error, and approximation error. - Approximation error includes truncation error from truncating infinite sums, and round-off error from representing numbers with a finite number of bits. - Relative and approximation errors are typically used to analyze errors in numerical solutions since the true value is unknown for real-world problems.

    Copyright:

    © All Rights Reserved
    Download as PPSX, PDF, TXT or read online from Scribd
    Download as ppsx, pdf, or txt
    140 views15 pages

    Lecture 3 - Errors in Numerical Computation

    Uploaded by

    Huzaika Matloob
    - Numerical methods yield approximate results that contain errors compared to the true analytical solution. - There are different types of errors including absolute error, relative error, and approximation error. - Approximation error includes truncation error from truncating infinite sums, and round-off error from representing numbers with a finite number of bits. - Relative and approximation errors are typically used to analyze errors in numerical solutions since the true value is unknown for real-world problems.

    Copyright:

    © All Rights Reserved
    Download as PPSX, PDF, TXT or read online from Scribd
    Download as ppsx, pdf, or txt
    of 15

    Errors in Numerical Computation

    Error
    • For many engineering problems, we cannot obtain analytical solutions.
    • Numerical methods yield approximate results, results that are close to the
    analytical solution.
    • If there's something that comes with an approximation, it's an error.
    Error
    •Analytical approach example:
    Find the root of f(x)=x−5.
    Analytical solution: f(x)=x−5=0, add +5 to both sides to get the answer x=5
    •Numerical solution:
    let's guess x=1: f(1)=1−5=−4. A negative number. Let's guess x=6:
    f(6)=6−5=1. A positive number. The answer must be between 1 and 6.
    •There are different types of errors
    •Absolute error
    •Relative Error
    •Approximation error
    Absolute Error
    • Absolute error or Absolute true error is a measure of how far 'off’ an estimated
    value is from a true value.
    Absolute Error = Actual Value - Estimated Value
    Relative Error
    • You first need to determine absolute error to calculate relative error. Relative
    error expresses how large the absolute error is compared with the total size of
    the object you are measuring. Relative error is expressed as a fraction or is
    multiplied by 100 and expressed as a percent.
    Relative Error = (Absolute Error / Known Value)*100
    Problem - Absolut Error and Relative Error

    • But the problem with absolute error and relative error is that we
    don't know what the exact true value is when we solve real life
    problems. The true value will be known only when we deal with
    functions that can be solved analytically (simple systems). So instead
    we use Approximation error.
    Approximation Error

    • When performing numerical computation, we use an iterative process. When


    performing these iterations, the approximate relative error as follows

    Ɛα = Present approximation - Previous approximation ⁄ Present approximation * 100

    can be used to determine if we are performing successively better computations.

    The signs of these errors can be either positive or negative dependent on the
    situation, so it is customary to take the absolute value of the error calculations as it’s
    only the magnitude of error that we are concerned with.
    Approximation Error
    • There are 2 types of approximation error:

    • Truncation error is the error made by truncating an infinite sum and


    approximating it by a finite sum.
    • Round-off error is the difference between an approximation of a number
    used in computation and its exact (correct) value.
    Truncation Error
    • For example, the infinite series 1/2 + 1/4 + 1/8 + 1/16 + 1/32 ... adds
    up to exactly 1.
    • However, if we truncate the series to only the first four terms, we get
    1/2 + 1/4 + 1/8 + 1/16 = 15/16, producing a truncation error of 1 -
    15/16, or 1/16.
    • Usually in evaluating logarithms, exponentials, trigonometric
    functions, hyperbolic functions etc., an infinite series is replaced by a
    finite series. Thus a truncation error is introduced in the computation.
    Error - Example
    Round-off Error
    • Roundoff error occurs because computers cannot represent some
    quantities exactly.
    • Round-off error occurs because computers use fixed number of bits
    and hence fixed number of binary digits to represent numbers.
    • Knowledge of roundoff error is important because they can lead to
    incorrect results.
    • When the number of bits required for representing a number are less
    then the number is usually rounded to fit the available number of
    bits. This is done either by chopping or by symmetric rounding.
    Round-off Error
    Chopping
    • Rounding a number by chopping amounts to dropping the extra digits. Here the given number is
    truncated. Suppose that we are using a computer with a fixed word length of four digits. Then the
    truncated representation of the number 72.32451 will be 72.32. The digits 451 will be dropped.
    Now to evaluate the error due to chopping let us consider the normalized representation of the
    given number i.e.

    • Therefore, chopping error is representing


    • So in general if a number x is the true value of a given number and fx * 10E is the normalized form
    of the rounded (chopped) number x and gx * 10 E-dis the normalized form of the chopping error then
    Round-off Error
    Symmetric Round-off Error
    The rounded rules are the following:

    4
    Round-off Error
    Symmetric Round-off Error
    • For example let x=72.918671 and y=18.63421 be two given numbers to be rounded to five
    digit numbers. The normalized form x and y are 0.7291867*10^2 and 0.1863421*10^2. On
    rounding these numbers to five digits we get 0.72919*10^2 and 0.18634*102 respectively.

    x = 72.918671 = 0.7291867*10^2
    = (0.72919+0.0000067)*10^2
    = (0.72919 + 0.67*10^-5)*10^2

    = (0.67 – 1) * 10^2-5
    Round-off Error
    Symmetric Round-off Error
    y = 18.63421 = 0.1863421*10^2
    = (0.18634+0.0000021)*10^2
    = (0.18634 + 0.21*10^-5)*10^2

    = (0.21) * 10^2-5

    You might also like