Part Four Curve Fitting • Chapter 14: Basic statistics and Linear regression • Chapter 15: General linear least-squares & Nonlinear regression • Chapter 16: Fourier analysis • Chapter 17: Polynomial interpolation • Chapter 18: Splines & Piecewise interpolation Chapter Objectives • Familiarizing yourself with some basic descriptive statistics and the normal distribution. • Knowing how to compute the slope and intercept of a best fit straight line with linear regression. • Knowing how to compute and understand the meaning of the coefficient of determination and the standard error of the estimate. • Understanding how to use transformations to linearize nonlinear equations so that they can be fit with linear regression. • Knowing how to implement linear regression with MATLAB. Statistics Review Measure of Location • Arithmetic mean: the sum of the individual data points (yi) divided by the number of points n: y= ∑ yi n • Median: the midpoint of a group of data. • Mode: the value that occurs most frequently in a group of data. Statistics Review Measures of Spread • Standard deviation: sy = St = ∑ ( yi − y )2 n −1 n −1 where St is the sum of the squares of the data residuals: St = ∑ (yi − y ) 2
and n-1 is referred to as the degrees of freedom.
( ) • Variance: ∑ (yi − y ) ∑ yi − ∑ yi / n 2 2 2
sy = 2 = n −1 n −1 • Coefficient of variation: sy c.v. = ×100% y Statistics Review • Example Table 6.495 6.565 6.625 6.435 6.635 6.395 6.655 6.655 6.665 6.595 6.515 6.715 6.625 6.485 6.775 6.605 6.755 6.505 6.615 6.555 6.575 6.715 6.555 6.685
• Solution ☞
=y ∑ 158.4 = y = 6.6 i = median 6.605 + 6.615 = 6.61 n 24 2
=sy 2 ∑ ( yi − y ) 2 0.217 = = 0.009435 n −1 23
= = ∑ yi 2 − (∑ yi ) 2 / n 1045.657 − 158.42 / 24 = 0.009435 n −1 23
=sy = St ∑ ( yi − y ) 2 = = s y 2 0.097133 n −1 n −1 sy 0.097133 c.v. = ×100% = ×100 =1.47% y 6.6 Normal Distribution Normal Distribution
modal class interval [ 6.6, 6.64 ]
6.6 + 6.64 = mode = 6.62 2
68% in [ y − s y , y + s y ] 95% in [ y − 2 s y , y + 2 s y ] = [6.405734, 6.794266 ] Descriptive Statistics in MATLAB • MATLAB has several built-in commands to compute and display descriptive statistics. Assuming some column vector s: – mean(s), median(s), mode(s) • Calculate the mean, median, and mode of s. mode is a part of the statistics toolbox. – min(s), max(s) • Calculate the minimum and maximum value in s. – var(s), std(s) • Calculate the variance and standard deviation of s • Note - if a matrix is given, the statistics will be returned for each column. Histograms in MATLAB • [n, x] = hist(s, x) – Determine the number of elements in each bin of data in s. x is a vector containing the center values of the bins. • [n, x] = hist(s, m) – Determine the number of elements in each bin of data in s using m bins. x will contain the centers of the bins. The default case is m=10 • hist(s, x) or hist(s, m) or hist(s) – With no output arguments, hist will actually produce a histogram. Histogram Example Linear Least-Squares Regression
y = a0 + a1 x + e = f ( x) + e
e = y − (a0 + a1 x) = y − f ( x) Linear Least-Squares Regression • Linear least-squares regression is a method to determine the “best” coefficients in a linear model for given data set. • “Best” for least-squares regression means minimizing the sum of the squares of the estimate residuals. For a straight line model, this gives: n n Sr = ∑ e = ∑ (yi − a0 − a1 xi ) 2 2 i i=1 i=1
• This method will yield a unique line for a given set
of data. Least-Squares Fit of a Straight Line • Using the model: y = a0 + a1 x
the slope and intercept producing the best fit
can be found using: n∑ xi yi − ∑ xi ∑ yi a1 = n∑ x − (∑ x ) 2 2 i i
a0 = y − a1 x Example V F n∑ xi yi − ∑ xi ∑ yi 8(312850) − (360)(5135) a1 = = = 19.47024 (m/s) (N) n∑ x − (∑ x ) 8(20400) − (360) 2 2 2 i i i xi yi (xi)2 x iy i a0 = y − a1 x = 641.875 −19.47024 (45) = −234.2857 1 10 25 100 250
2 20 70 400 1400
3 30 380 900 11400 Fest = −234.2857 +19.47024v
4 40 550 1600 22000
5 50 610 2500 30500
6 60 1220 3600 73200
7 70 830 4900 58100
8 80 1450 6400 116000
Σ 360 5135 20400 312850
Quantification of Error • Recall for a straight line, the sum of the squares of the estimate residuals: n n Sr = ∑ e = ∑ (yi − a0 − a1 xi ) 2 2 i i=1 i=1
St = ∑ i ( y − y )2
• Standard error of the estimate:
Sr St s y/ x = sy = n−2 n −1 Standard Error of the Estimate • Regression data showing (a) the spread of data around the mean of the dependent data and (b) the spread of the data around the best fit line:
• The reduction in spread represents the improvement due to
linear regression. Coefficient of Determination • The coefficient of determination r2 is the difference between the sum of the squares of the data residuals and the sum of the squares of the estimate residuals, normalized by the sum of the squares of the data residuals: St − Sr r = 2
St • r2 represents the percentage of the original uncertainty explained by the model. • For a perfect fit, Sr=0 and r2=1. • If r2=0, there is no improvement over simply picking the mean. • If r2<0, the model is worse than simply picking the mean! Example V F (m/s) (N) Fest = −234.2857 +19.47024v St = ∑ (yi − y ) = 1808297 2 i xi yi a0+a1xi (yi- ȳ)2 (yi-a0-a1xi)2
Sr = ∑ (yi − a0 − a1 xi ) = 216118 1 10 25 -39.58 380535 4171 2
2 20 70 155.12 327041 7245
1808297 3 30 380 349.82 68579 911 sy = = 508.26 8 −1 4 40 550 544.52 8441 30 216118 s y/ x = = 189.79 5 50 610 739.23 1016 16699 8−2 6 60 1220 933.93 334229 81837 1808297 − 216118 r2 = = 0.8805 7 70 830 1128.63 35391 89180 1808297 8 80 1450 1323.33 653066 16044 88.05% of the original uncertainty has been explained by the Σ 360 5135 1808297 216118 linear model Nonlinear Relationships • Linear regression is predicated on the fact that the relationship between the dependent and independent variables is linear - this is not always the case. • Three common examples are: exponential : y = α1eβ1 x
power : y = α2 x β2 x saturation - growth - rate : y = α 3 β3 + x Linearization of Nonlinear Relationships • One option for finding the coefficients for a nonlinear fit is to linearize it. For the three common models, this may involve taking logarithms or inversion: Model Nonlinear Linearized
exponential : y = α1eβ1 x ln y = ln α1 + β1 x
power : y = α2 x β2 log y = log α 2 + β2 log x
x 1 1 β3 1 saturation - growth - rate : y = α 3 = + β3 + x y α3 α3 x Transformation Examples Transformation Examples
• Solution ☞ =y α 2 x β2 = log y log α 2 + β 2 log x
12.606 20.515 = = 1.57657 X = Y = 2.5644 8 8 n∑ xi yi − ∑ xi ∑ yi 8(33.622) − 12.606(20.515) =a1 = = 1.9842 n∑ xi − (∑ xi ) − 2 2 2 8(20.516) (12.606) a0 =− y a1 x = 2.5644 − 1.9842(1.5757) =− 0.5620
log y = −0.5620 + 1.9842 log x
−0.562 1.9842 = = F 10 v 0.274v1.9842 Transformation Examples
1,600 Data ☞ • Solution 2,000 Data 1,200 Regression line 1,600 Regression line
1,200 800 F [N]
F [N] 800 400 =F 19.470 v − 234.286 400 F = 0.274 v1.9842 0 0 0 20 40 60 80 100 0 20 40 60 80 100 -400 -400 v [m/s] v [m/s] Linear Regression Program MATLAB Functions • MATLAB has a built-in function polyfit that fits a least-squares nth order polynomial to data: – p = polyfit(x, y, n) • x: independent data • y: dependent data • n: order of polynomial to fit • p: coefficients of polynomial f(x)=p1xn+p2xn-1+…+pnx+pn+1 • MATLAB’s polyval command can be used to compute a value using the coefficients. – y = polyval(p, x)
