Curve Fitting - F22

11/3/2022
Public Works Department

Master Degree
Dr. Asmaa Moddather

Soil Mechanics and Foundations
Faculty of Engineering – Cairo University
FALL 2022
10
11
1
11/3/2022
Introduction
Data is often given for discrete values along a

continuum.
However, you may require estimates at points between

the discrete values.
The present part describes techniques to fit curves to

such data to obtain intermediate estimates.
In addition, you may require a simplified version of a

complicated function.
Dr. Asmaa Moddather – Numerical Analysis – Fall 2022
12
Introduction
One way to do this is to compute values of the function

at a number of discrete values along the range of
interest. Then, a simpler function may be derived to fit
these values.
Both of these applications are known as curve fitting.
There are two general approaches for curve fitting that

are distinguished from each other on the basis of the
amount of error associated with the data.
13
2
11/3/2022
Introduction
First, where the data exhibits a significant degree of error or
“noise”, the strategy is to derive a single curve that represents
the general trend of the data.
Because any individual data
point may be incorrect, we make
no effort to intersect every point.
Rather, the curve is designed
to follow the pattern of the
points taken as a group. One
approach of this nature is called
least-squares regression.
14
Introduction
Second, where the data is known to be
very precise, the basic approach is to
fit a curve or a series of curves that
pass directly through each of the
points. Such data usually originates
from tables. Examples are values for the
density of water or for the heat capacity
of gases as a function of temperature.
The estimation of values between well-

known discrete points is called
interpolation. Dr. Asmaa Moddather – Numerical Analysis – Fall 2022
15
3
11/3/2022
Noncomputer methods
The simplest method for fitting a curve to data is to plot the

points and then sketch a line that visually conforms to the data.
Although this is a valid option when quick estimates are

required, the results are dependent on the subjective viewpoint
of the person sketching the curve.
16
Engineering Practice
Two types of applications are generally encountered when

fitting experimental data: trend analysis and hypothesis
testing.
Trend analysis represents the process of using the pattern

of the data to make predictions. For cases where the data is
measured with high precision, you might utilize
interpolating polynomials. Imprecise data is often analyzed
with least-squares regression.
17
4
11/3/2022
Trend analysis may be used to predict or forecast values of

the dependent variable. This can involve extrapolation
beyond the limits of the observed data or interpolation
within the range of the data.
hypothesis testing: an existing mathematical model is

compared with measured data. If the model coefficients
are unknown, it may be necessary to determine values that
best fit the observed data.
18
On the other hand, if estimates of the model coefficients

are already available, it may be appropriate to compare
predicted values of the model with observed values to test
the adequacy of the model.
Often, alternative models are compared and the “best” one

is selected on the basis of empirical observations.
19
5
11/3/2022
Curve fitting is important in other numerical methods

such as integration and the approximate solution of
differential equations.
Finally, curve-fitting techniques can be used to derive

simple functions to approximate complicated functions.
20
Curve Fitting
Least Square Regression
Interpolation
Fourier Approximation
21
6
11/3/2022
Linear Regression
Multiple Regression
General linear least squares
Nonlinear Regression
22
Where substantial error is associated with data, polynomial

interpolation is inappropriate and may yield unsatisfactory
results when used to predict intermediate values.
23
7
11/3/2022
Linear Regression
The simplest example of a least-squares approximation is fitting

a straight line to a set of paired observations: (x1, y1), (x2, y2), . . . ,
(xn, yn). The mathematical expression for the straight line is:
24
Linear Regression
25
8
11/3/2022
Linear Regression
26
Linear Regression
27
9
11/3/2022
Linear Regression
We need a criteria to fit

data !!
e=
Minimum error
for all points
One best solution
28
Linear Regression
One strategy for fitting a “best” line through the data would be to
minimize the sum of the residual errors for all the available data:
-e
+e
Σe = zero !!
29
10
11/3/2022
Linear Regression
more than one solution !!
30
Linear Regression
The minimax criterion:
The line is chosen that

minimizes the maximum
distance that an individual
point falls from the line.
31
11
11/3/2022
Linear Regression
How to calculate ao and a1?
This criterion has a number of advantages, overcomes the
shortcomings of the aforementioned approaches and the fact
that it yields a unique line for a given set of data.
32
Linear Regression
Differentiate Sr with respect to each coefficient:
If these derivatives = zero  minimum Sr
33
12
11/3/2022
Linear Regression
34
Linear Regression
Fit a straight line to the x and y values in the following
Table:
35
13
11/3/2022
Linear Regression
36
Linear Regression
xi yi x2i xiyi
1 0.5 1 0.5
2 2.5 4 5
3 2 9 6
4 4 16 16
5 3.5 25 17.5
6 6 36 36
7 5.5 49 38.5
28 24 140 119.5
37
14
11/3/2022
Linear Regression
Quantification of data spread around regression line:
Sy/x : The standard error of estimates 
38
Linear Regression
Quantification of Error reduction:
r2: Coefficient of determination 
r: correlation coefficient
39
15
11/3/2022
Linear Regression
40
Linear Regression
7
6 y = 0.8393x + 0.0714
r² = 0.8683
5
4
y
3
2
1
0
0 2 4 6 8
x
41
16
11/3/2022
Linear Regression
Always inspect the plot of

points along with your
regression curve !!
42
Linearization of Nonlinear Relationships
Linear regression provides a powerful technique for fitting

a best line to data.
This is not always the case, and the first step in any
regression analysis should be to plot and visually inspect
the data to ascertain whether a linear model applies.
Some data are curvilinear. In some cases, techniques such

as polynomial regression are appropriate.
For others, transformations can be used to express the data

in a form that is compatible with linear regression.
43
17
11/3/2022
Exponential equation
44
Power equation
45
18
11/3/2022
Saturation growth rate equation
46
Linear Regression
Fit a power equation to the x and y values in the following
Table:
47
19
11/3/2022
Linear Regression
New x and y  use linear regression

a1 = β2
a0 = log α2
48
Linear Regression
49
20
11/3/2022
Polynomial Regression
Convert into
matrix form
50
51
21
11/3/2022
Fit a second-order polynomial to the data in the
following Table:
52
53
22
11/3/2022
54
55
23
11/3/2022
56
Multiple Linear Regression
A useful extension of linear regression is the case where y is a

linear function of two or more independent variables:
Such an equation is particularly useful when fitting experimental

data where the variable being studied is often a function of two
other variables.
57
24
11/3/2022
For this two-dimensional case, the regression “line” becomes a

“plane”
58
59
25
11/3/2022
Use multiple linear regression to fit this data:
60
61
26
11/3/2022
62
Interpolation
Newton Polynomial
Lagrange Polynomial
Polynomial Coefficient
Inverse Interpolation
Splines
Multidimensional Interpolation
63
27
11/3/2022
Interpolation
You will frequently have occasion to estimate intermediate

values between precise data points.
The most common method used for this purpose is

polynomial interpolation with the the general formula for
an nth-order polynomial is for n + 1 data points.
There is one and only one polynomial of order n that passes

through all the points.
64
Interpolation
There is only one straight line
(first-order polynomial) that
connects two points.
Only one parabola (second

order polynomial) connects a
set of three points.
Only one cubic (third order

polynomial) connects a set of
four points.
65
28
11/3/2022
Newton Polynomial
The simplest form of interpolation

is to connect two data points with a
straight line. This technique, called
linear interpolation, is depicted
graphically, using similar triangles:
66
Newton Polynomial
Estimate the natural logarithm of 2 using linear

interpolation.
If ln 1 = 0 and ln 6 = 1.791759.
If ln 1 = 0 and ln 4 =1.386294.
(ln 2 = 0.6931472)
67
29
11/3/2022
Newton Polynomial
68
Newton Polynomial
69
30
11/3/2022
Newton Polynomial
The error in previous Example resulted from our

approximating a curve with a straight line.
A strategy for improving the estimate is to introduce some

curvature into the line connecting the points.
If three data points are available, this can be accomplished

with a second-order polynomial (also called a quadratic
polynomial or a parabola):
70
Newton Polynomial
71
31
11/3/2022
Newton Polynomial
72
Newton Polynomial
Estimate the natural logarithm of 2 using quadratic

interpolation.
If ln 1 = 0, ln 4 = 1.386294, and ln 6 = 1.791759.
(ln 2 = 0.6931472)
73
32
11/3/2022
Newton Polynomial
74
Newton Polynomial
75
33
11/3/2022
Newton Polynomial
76
Newton’s divided-difference interpolating

polynomial
77
34
11/3/2022
Newton’s divided-difference interpolating

polynomial
78
Newton Polynomial
Estimate the natural logarithm of 2 using quadratic

interpolation.
If xo = 1, x1 = 4, x2 = 6, x3 = 5
ln 5 = 1.609438.
(ln 2 = 0.6931472)
79
35
11/3/2022
Newton Polynomial
80
Newton Polynomial
81
36
11/3/2022
Newton Polynomial
82
Lagrange Polynomial
The Lagrange interpolating polynomial is simply a

reformulation of the Newton polynomial that avoids the
computation of divided differences. It can be represented:
83
37
11/3/2022
Lagrange Polynomial
84
Lagrange Polynomial
Use a Lagrange interpolating polynomial of the first and

second order to evaluate ln 2 on the basis of the following
data:
85
38
11/3/2022
Lagrange Polynomial
86
Verification Using Excel
Excel Trendline
The data analysis Toolpack
87
39
11/3/2022
Trendline
88
Trendline
89
40
11/3/2022
Toolpack

90
Toolpack
91
41
11/3/2022
Toolpack
92
Toolpack
93
42

Curve Fitting - F22

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Curve Fitting - F22

Uploaded by

Copyright:

Available Formats

11/3/2022

Public Works Department

Dr. Asmaa Moddather

Data is often given for discrete values along a

However, you may require estimates at points between

The present part describes techniques to fit curves to

In addition, you may require a simplified version of a

One way to do this is to compute values of the function

Both of these applications are known as curve fitting.

There are two general approaches for curve fitting that

The estimation of values between well-

The simplest method for fitting a curve to data is to plot the

Although this is a valid option when quick estimates are

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Two types of applications are generally encountered when

Trend analysis represents the process of using the pattern

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Trend analysis may be used to predict or forecast values of

hypothesis testing: an existing mathematical model is

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

On the other hand, if estimates of the model coefficients

Often, alternative models are compared and the “best” one

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Curve fitting is important in other numerical methods

Finally, curve-fitting techniques can be used to derive

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Least Square Regression

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Least Square Regression

General linear least squares

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Least Square Regression

Where substantial error is associated with data, polynomial

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

The simplest example of a least-squares approximation is fitting

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

We need a criteria to fit

for all points

One best solution

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

more than one solution !!

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

The minimax criterion:

The line is chosen that

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

This criterion has a number of advantages, overcomes the

shortcomings of the aforementioned approaches and the fact

that it yields a unique line for a given set of data.

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Differentiate Sr with respect to each coefficient:

If these derivatives = zero  minimum Sr

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Fit a straight line to the x and y values in the following

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022

Sy/x : The standard error of estimates 

Dr. Asmaa Moddather – Numerical Analysis – Fall 2022