Scribd Logo
Upload

Welcome to Scribd!

  • 41 views

    Chapter 8: Performance Surfaces and Optimum Points: Brandon Morgan 1/14/2021

    Uploaded by

    samaneh keshavarzi
    This document discusses using Taylor series approximations to estimate a scalar function F(x). It provides the first and second derivatives of F(x) and uses them to develop second-order Taylor series approximations around two points: x* = -0.5 and x* = 1.1. Plots show that the approximation is accurate near x* = -0.5 but not near x* = 1.1, since the original function is undefined at that point.

    Copyright:

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

    Chapter 8: Performance Surfaces and Optimum Points: Brandon Morgan 1/14/2021

    Uploaded by

    samaneh keshavarzi
    41 views4 pages
    This document discusses using Taylor series approximations to estimate a scalar function F(x). It provides the first and second derivatives of F(x) and uses them to develop second-order Taylor series approximations around two points: x* = -0.5 and x* = 1.1. Plots show that the approximation is accurate near x* = -0.5 but not near x* = 1.1, since the original function is undefined at that point.

    Original Title

    E8_1

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses using Taylor series approximations to estimate a scalar function F(x). It provides the first and second derivatives of F(x) and uses them to develop second-order Taylor series approximations around two points: x* = -0.5 and x* = 1.1. Plots show that the approximation is accurate near x* = -0.5 but not near x* = 1.1, since the original function is undefined at that point.

    Copyright:

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

    Chapter 8: Performance Surfaces and Optimum Points: Brandon Morgan 1/14/2021

    Uploaded by

    samaneh keshavarzi
    This document discusses using Taylor series approximations to estimate a scalar function F(x). It provides the first and second derivatives of F(x) and uses them to develop second-order Taylor series approximations around two points: x* = -0.5 and x* = 1.1. Plots show that the approximation is accurate near x* = -0.5 but not near x* = 1.1, since the original function is undefined at that point.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 4

    Chapter 8: Performance Surfaces and Optimum Points

    Brandon Morgan

    1/14/2021

    E8.1
    We are given the following scalar function

    1
    F (x) =
    x3 − 0.75x − 0.5
    The second-order Taylor series approximation for a scalar function, F (x), is given by:

    d 1 d2
    F (x) = F (x∗ ) + F (x)|x=x∗ (x − x∗ ) + F (x)|x=x∗ (x − x∗ )2
    dx 2 dx2
    Here we get the following first and second derivatives of the function:

    3x2 − 0.75
    F 0 (x) = −
    (x3 − 0.75x − 0.5)2

    2(−6x4 + 2.25x2 − 1.5x − 0.5625)


    F 00 (x) = −
    (x3 − 0.75x − 0.5)3

    Thus, we end up with the following expansion:

    1 3(x∗ )2 − 0.75 ∗ 1 2(−6(x∗ )4 + 2.25(x∗ )2 − 1.5(x∗ ) − 0.5625)


    F (x) = − (x−(x ))− (x−(x∗ ))2
    (x∗ )3 − 0.75(x∗ ) − 0.5 ((x∗ )3 − 0.75(x∗ ) − 0.5)2 2 ((x∗ )3 − 0.75(x∗ ) − 0.5)3

    When x∗ = −0.5, our Taylor expansion reduces down to:

    F (x) = −4 + 0(x + 0.5) + 24(x + 0.5)2 = −4 + 24(x + 0.5)2

    When x∗ = 1.1, our Taylor expansion reduces down to:

    F (x) = 166.67 − 80000(x − 1.1) + 3.83 ∗ 107 (x − 1.1)2

    1
    3
    We can now plot our scalar function and its second order Taylor expansion at each point to obersve its
    accuracy.

    library(ggplot2)

    ## Warning: package ’ggplot2’ was built under R version 3.6.3

    fun = function(x) { # our original function


    1/(x^3-0.75*x-0.5)
    }
    secondDegree1 = function(x) { # taylor expansion at -0.5
    -4+24*(x+0.5)^2
    }
    secondDegree2 = function(x) { # taylor expansion at 1.1
    166.67-80000*(x-1.1)+3.83*(10^7)*(x-1.1)^2
    }

    Second-order term aat x = −0.5

    ggplot(data.frame(x=c(-1, 0)), aes(x=x))+stat_function(fun=fun, color = "blue")+


    stat_function(fun=secondDegree1, color = "red")+
    scale_y_continuous(limits = c(-7,7))+geom_hline(yintercept=0)+geom_vline(xintercept = 0)

    0
    y

    −4

    −1.00 −0.75 −0.50 −0.25 0.00


    x

    2
    As we can see from the plot above, our second order Taylor approximation does a very good job of approxi-
    mating our function at the point x = −0.5 within a small interval.

    Second-order term at x = 1.1

    ggplot(data.frame(x=c(0.5, 1.5)), aes(x=x))+stat_function(fun=fun, color = "blue")+geom_hline(yintercept

    −40
    y

    −80

    −120

    0.0 0.5 1.0 1.5


    x

    ggplot(data.frame(x=c(0.5, 1.5)), aes(x=x))+stat_function(fun=secondDegree2, color = "red")+geom_hline(y

    3
    4e+07

    3e+07
    y

    2e+07

    1e+07

    0e+00

    0.0 0.5 1.0 1.5


    x

    Unfortunately, I was unable to plot both functions on top of each other; however, we can see that our function
    approaches infinity as x → 1.1 from both sides. Because this point is undefined, our Taylor approximation
    does a poor job at approximating this point, as one can tell from its graph.

    You might also like