Lecturer, School of Civil & Environmental Engineering. Members of Road and Transport Engineering Chair Addis Ababa Institute of Technology Addis Ababa University masteshumet@gmail.com mastewal.shumet@aait.edu.et mastewal.shumet@aau.edu.et 12/4/2024 Mastewal S. 1 Numerical Methods (CENG-3081) Chapter 5 CURVE FITTING
12/4/2024 Mastewal S. 2 Introduction To Curve Fitting • Curve fitting is a fundamental concept in numerical methods where a curve is constructed to best fit a set of data points. • The goal of curve fitting is to create a mathematical model that approximates the relationship between the variables in the data. Data Points : - • These are the given set of points (𝑥𝑖 , 𝑦𝑖 ), where 𝑥𝑖 is the independent variable, and 𝑦𝑖 is the dependent variable. The task is to find a curve that passes through or comes close to these points. 12/4/2024 Mastewal S. 3 CURVE FITTING • Describes techniques to fit curves (curve fitting) to discrete data to obtain intermediate estimates.
• There are two general approaches to curve fitting:
– Data exhibit a significant degree of scatter. The strategy is to derive a single curve that represents the general trend of the data. – Data is very precise. The strategy is to pass a curve or a series of curves through each of the points.
• In engineering two types of applications are encountered:
– Trend analysis. Predicting values of dependent variable, may include extrapolation beyond data points or interpolation between data points. – Hypothesis testing. Comparing existing mathematical model with measured data.
12/4/2024 Mastewal S. 4 12/4/2024 Mastewal S. 5 Mathematical Background Simple Statistics/ • In course of engineering study, if several measurements are made of a particular quantity, additional insight can be gained by summarizing the data in one or more well chosen statistics that convey as much information as possible about specific characteristics of the data set. • These descriptive statistics are most often selected to represent – The location of the center of the distribution of the data, – The degree of spread of the data.
12/4/2024 Mastewal S. 6 • Arithmetic mean. The sum of the individual data points (yi) divided by the number of points (n).
y= y i
n i = 1, , n • Standard deviation. The most common measure of a spread for a sample.
y − ( y ) 2 2 St /n Sy = n −1 or S 2 = i i
n −1 y
S t = ( yi − y ) 2
12/4/2024 Mastewal S. 7 • Variance. Representation of spread by the square of the standard deviation.
S y2 = i ( y − y ) 2
Degrees of freedom n −1
• Coefficient of variation. Has the utility to quantify the
spread of data.
Sy c.v. = 100% y
12/4/2024 Mastewal S. 8 Least Squares Regression
Linear Regression • Fitting a straight line to a set of paired observations: (x1, y1), (x2, y2),…,(xn, yn). y=a0+a1x+e a1- slope a0- intercept e- error, or residual, between the model and the observations
12/4/2024 Mastewal S. 9 Criteria for a “Best” Fit/ • Minimize the sum of the residual errors for all available data: n n
e = (y i =1 i i =1 i − ao − a1 xi )
n = total number of points
• However, this is an inadequate criterion, so is the sum of the absolute values n n
e = y i =1 i i =1 i − a0 − a1 xi
12/4/2024 Mastewal S. 10 • Best strategy is to minimize the sum of the squares of the residuals between the measured y and the y calculated with the linear model: n n n S r = ei2 = ( yi , measured − yi , model ) 2 = ( yi − a0 − a1 xi ) 2 i =1 i =1 i =1
• Yields a unique line for a given set of data.
12/4/2024 Mastewal S. 11 Least-Squares Fit of a Straight Line/ S r = −2 ( yi − ao − a1 xi ) = 0 ao S r = −2 ( yi − ao − a1 xi ) xi = 0 a1 0 = yi − a 0 − a1 xi 0 = yi xi − a 0 xi − a1 xi2
a 0 = na0 Normal equations, can be
na0 + ( xi )a1 = yi solved simultaneously
n xi yi − xi yi a1 = n x − ( xi ) 2 2 i Mean values a0 = y − a1 x 12/4/2024 Mastewal S. 12 12/4/2024 Mastewal S. 13 “Goodness” of our fit/ If • Total sum of the squares around the mean for the dependent variable, y, is St • Sum of the squares of residuals around the regression line is Sr • St-Sr quantifies the improvement or error reduction due to describing data in terms of a straight line rather than as an average value. St − S r r =2
St r2-coefficient of determination Sqrt(r2) – correlation coefficient 12/4/2024 Mastewal S. 14 • For a perfect fit Sr=0 and r=r2=1, signifying that the line explains 100 percent of the variability of the data. • For r=r2=0, Sr=St, the fit represents no improvement.
12/4/2024 Mastewal S. 15 Polynomial Regression • Some engineering data is poorly represented by a straight line. • For these cases a curve is better suited to fit the data. • The least squares method can readily be extended to fit the data to higher-order polynomials.
12/4/2024 Mastewal S. 16 General Linear Least Squares y = a0 z0 + a1 z1 + a2 z 2 + + am zm + e z0 , z1, , zm are m + 1 basis functions Y = Z A+ E Z − matrix of the calculated values of the basis functions at the measured values of the independent variable Y− observed valued of the dependent variable A− unknown coefficients E− residuals 2 Minimized by taking its partial n m S r = yi − a j z ji derivative w.r.t. each of the coefficients and setting the i =1 j =0 resulting equation equal to zero 12/4/2024 Mastewal S. 17 Example
12/4/2024 Mastewal S. 18 Solution
12/4/2024 Mastewal S. 19 Linearization of Nonlinear Relationships
12/4/2024 Mastewal S. 20 Linearization of Nonlinear Relationships
12/4/2024 Mastewal S. 21 Linearization of a Power Equation (Example)
-0.300 1.752 0.501 12/4/2024 Mastewal S. 23 MULTIPLE LINEAR REGRESSION
12/4/2024 Mastewal S. 24 Determine Multi linear Equation Constants 𝒚 = 𝒂𝒐 + 𝒂𝟏 𝒙𝟏 + 𝒂𝟐 𝒙𝟐
𝑥1 𝑥2 y 0 0 5 2 1 10 2.5 2 9 1 3 0 4 6 3 7 2 27
12/4/2024 Mastewal S. 25 Interpolation • Used to estimate intermediate values between precise data points. • The most common method used is the polynomial interpolation. • For n+1 data points, there is a unique polynomial of order n that passes through all the points. 𝑓(𝑥) = 𝑎𝑜 + 𝑎1 𝑥 + 𝑎2 𝑥2 + … + 𝑎𝑛 𝑥𝑛 • Newton and Lagrange interpolating polynomials are the most widely used methods
12/4/2024 Mastewal S. 26 Newton's Divided Difference Interpolating Polynomials • Called linear interpolation (similar triangles) (linear) f1 ( x) − f ( xo ) f ( x1 ) − f ( xo ) = x − xo x1 − xo f ( x1 ) − f ( xo ) f1 ( x) = f ( xo ) + ( x − xo ) x1 − xo f1 ( x) = bo + b1 ( x − xo )
bo = f ( xo )
f ( x1 ) − f ( xo ) b1 = x1 − xo
12/4/2024 Mastewal S. 27 Cont’d • The smaller the interval the better approximation • Using Curvature approximation is better than that of straight line estimate Second order ( Quadratic Interpolating) polynomial 𝑓2(𝑥) = 𝑏𝑜 + 𝑏1(𝑥 − 𝑥𝑜) + 𝑏2(𝑥 − 𝑥0) (𝑥 − 𝑥1) 𝒃𝟎 Can be solved by setting → 𝒙 = 𝒙𝟎 𝒃𝟏 Can be solved by setting → 𝒙 = 𝒙𝟏 𝒃𝟐 Can be solved by setting → 𝒙 = 𝒙𝟐
f ( x1 ) − f ( xo ) 𝑏𝑜 = 𝑓2 𝑥𝑜 b1 = x1 − xo f ( x2 ) − f ( x1 ) f ( x1 ) − f ( xo ) − x2 − x1 x1 − xo b2 = 12/4/2024 x2 −S.xo Mastewal 28 General form of Newton's Interpolating Polynomials • The generalized to fit an nth order polynomial to n+1 data points. The nth order polynomial is 𝒇𝒏 𝒙 = 𝒃𝟎 + 𝒃𝟏 𝒙 − 𝒙𝒐 + 𝒃𝟐 𝒙 − 𝒙𝒐 𝒙 − 𝒙𝟏 +. . + 𝒃𝒏(𝒙 − 𝒙𝒐)(𝒙 − 𝒙𝟏) … . (𝒙 − 𝒙𝒏 − 𝟏)
𝑇ℎ𝑒 𝑑𝑎𝑡𝑎 𝑝𝑜𝑖𝑛𝑡𝑠 𝑐𝑎𝑛 𝑏𝑒 𝑢𝑠𝑒𝑑 𝑡𝑜 𝑒𝑣𝑎𝑙𝑢𝑎𝑡𝑒 𝑡ℎ𝑒
𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡𝑠 𝑏𝑜, 𝑏1, 𝑏2, … . , 𝑏𝑛.
𝑏0 = 𝑓(𝑥𝑜) 𝑏1 = 𝑓[𝑥1, 𝑥𝑜] 𝑏2 = 𝑓[𝑥2, 𝑥1, 𝑥𝑜] . . . 12/4/2024 𝑏𝑛 = 𝑓[𝑥𝑛,Mastewal 𝑥𝑛 S.− 1, … , 𝑥1, 𝑥𝑜] 29 Cont’d • First finite divided difference is given by f ( xi ) − f ( x j ) f [ xi , x j ] = xi − x j • The second finite divided difference is expressed by f [ xi , x j ] − f [ x j , xk ] f [ xi , x j , xk ] = xi − xk • Similarly the 𝑛𝑡ℎ finite divided difference is
f [ xn , xn −1 ,..., x1 ] − f [ xn −1 , xn − 2 ,..., xo ] f [ xn , xn −1 ,..., x1, xo ] = xn − xo
12/4/2024 Mastewal S. 30 Example (First Order Polynomial) • The upward velocity of a rocket is given as a function of time in table below • Velocity as a function of time. 𝑚 𝑡 (𝑠) 𝑉 ( ) 𝑠 0 0 10 227.04 15 362.78 20 517.35 22.5 602.67 30 901.67 • Determine the value of the velocity at 𝑡 = 16 seconds using first order polynomial interpolation by Newton’s divided difference polynomial method. 12/4/2024 Mastewal S. 31 Solution • For linear interpolation, the velocity is given by 𝑣 𝑡 = 𝑏0 + 𝑏1 (𝑡 − 𝑡0 ) • Since we want to find the velocity at 𝑡 = 16 , and we are using a first order polynomial, we need to choose the two data points that are closest to 𝑡 = 16 that also bracket 𝑡 = 16 to evaluate it. • The two points are 𝑡 = 15 and 𝑡 = 20. • Then 𝑡0 = 15, 𝑣 𝑡0 = 362.78 𝑡1 = 20, 𝑣 𝑡1 = 517.35 𝑔𝑖𝑣𝑒𝑠 𝑏0 = 𝑣 𝑡0 = 362.78 𝑣 𝑡1 − 𝑣 𝑡0 517.35 − 362.78 𝑏1 = = = 30.914 𝑡1 − 𝑡0 20 − 15
𝒗 𝟏𝟔 = 𝟑𝟔𝟐. 𝟕𝟖 + 𝟑𝟎. 𝟗𝟏𝟒
12/4/2024 𝟏𝟔 − 𝟏𝟓 = 𝟑𝟗𝟑. 𝟔𝟗 𝒎/𝒔 Mastewal S. 32 Example (Quadratic Polynomial) • The upward velocity of a rocket is given as a function of time in table below • Velocity as a function of time. 𝑡 (𝑠) 𝑚 𝑉 ( ) 𝑠 0 0 10 227.04 15 362.78 20 517.35 22.5 602.67 30 901.67 • Determine the value of the velocity at 𝑡 = 16 seconds using second order polynomial interpolation by Newton’s divided difference polynomial method. 12/4/2024 Mastewal S. 33 Solution • For Quadratic interpolation, the velocity is given by 𝑣 𝑡 = 𝑏0 + 𝑏1 𝑡 − 𝑡0 + 𝑏2 𝑡 − 𝑡0 𝑡 − 𝑡1 • Since we want to find the velocity at 𝑡 = 16 , and we are using a second order polynomial, we need to choose the three data points that are closest to 𝑡 = 16 that also bracket 𝑡 = 16 to evaluate it. • The three points are 𝑡 = 10, t = 15 and 𝑡 = 20. • Then 𝑡0 = 10, 𝑣 𝑡0 = 227.04 𝑡1 = 15, 𝑣 𝑡1 = 362.78 𝑡2 = 20, 𝑣 𝑡2 = 517.35 𝑔𝑖𝑣𝑒𝑠 𝑏0 = 𝑣 𝑡0 = 227.04
