Compute the behavior of reality even if it is impossible to observe the processes (for example a black hole in astrophysics).

Size: px
Start display at page:

Download "Compute the behavior of reality even if it is impossible to observe the processes (for example a black hole in astrophysics)."

Transcription

1 1 Introduction Read sections 1.1, , 1.2.6, 1.3.8, 1.3.9, 1.4. Review questions , , The subject of Scientific Computing is to simulate the reality. Simulation is the representation and emulation of a physical system or a physical process with the aid of a computer (the computational model). Hence, Scientific Computing is an integration of mathematics (numerical analysis), computer science, and engineering sciences. What can be done if we have a computational model of a process? Compute the behavior of reality even if it is impossible to observe the processes (for example a black hole in astrophysics). Make experiments considerably faster and cheaper (for example the development of integrated circuits, simulated crash tests for cars). The whole process of the solution of an applied problem can be subdivided into a number of subproblems: 1. Develop a mathematical model for the system under consideration. This consists most often of a system of equations and inequalities. 2. Develop algorithms for the numerical solution of the equations describing the mathematical model. This is conveniently considered being the numerical model. 3. Implement the resulting algorithms in a computer program. This should be an equivalent description of the numerical model. 4. Run the program on a computer. Because of compiler and hardware dependencies (for instance rounding errors), the results may differ from what is expected from the numerical model. 5. Represent the results in a form which is easily comprehensible. This is often done by visualizing the results graphically. 6. Interpret and validate the results. If necessary, repeat some or all of the previous steps. The first step is usually called mathematical modeling. Even if it is called mathematical, it is only the language which is mathematical. Indeed, mathematical modeling is a part of the engineering science which deals with the process under consideration. But it really important that the scientist or engineer who will develop the numerical methods takes part in this process. Usually there are many different ways of describing a process, but their ability to be efficiently approximated numerically may differ considerably. Numerical analysis consists of the steps two and three of the solution process. This is what we consider during the lectures. 1

2 The general strategy for the solution of a mathematical model consists of a replacement of harder problems by easier ones which are solvable on a computer and whose solution comes close to that of the original hard problem: Replace an infinite process by a finite one. Replace differential expressions by algebraic ones. Replace nonlinear models by linear ones. After the replacement we obtain only an approximate problem such that the solution is also only an approximate one. This is not necessarily a big problem because the complete solution process (starting with the mathematical formulation and ending with the solution representation) includes already a number of approximations: Before the real computation starts: model errors; measurement errors; results of previous computations which will be used in the present model. During computations: rounding errors; discretization errors; truncation errors during iterative processes. The accuracy of the final result depends on all these sources of errors: (i) The uncertainty in the input data may be amplified because of the properties of the mathematical model. (ii) Perturbations during the computations may be amplified because of the properties of the algorithm. Example 1.1. We want to compute the surface of the earth by using the formula Among the approximations are, for example: The earth is considered to be a sphere. A 4πr 2 The value for the radius is obtained by measurements and previous computations. The value for π requires the truncation of an infinite process. 2

3 The input data as well as the numerical computations lead to rounding errors. The problem can even become much harder if one tries to apply numerical algorithms. Example 1.2. The following identity holds true: 1 a b a c a b a c b We take for example a 1000, b 0 001, and c If we evaluate both expressions on a small hand-held computer with eight digits accuracy, then we obtain the following results: Using the left expression: 0 0. Using the right expression: Which result is more correct? The problem is that different algorithms for identical mathematical models may lead to large errors. For those of you who are curios: In MATLAB both expressions yield because of the double precision arithmetic used. Example 1.3. Consider the following expression: W b 6 a 2 11a 2 b 2 b 6 121b b 8 a This formula was implemented on an IBM 4381 computer using the Fortran programming language and the VS Fortran system in different precisions. Taking the input data a and b , the programs provided the following results: precision rounding unit result single double extended b The exact result is W Absolute And Relative Errors In order to be able to provide some estimations of the accuracy of the input data and the results of certain computations, it is usual to use the notion of the absolute and relative errors of a number, a vector, a matrix, and so on. The general rules are: absolute error approximate value true value relative error absolute error true value 3

4 The last equation can also be rewritten in the more convenient form approximate value true value 1 relative error Because the true value is not available, it is only possible to estimate, or bound, the errors. It is common to use often statements like three digits of accuracy (which amounts to a bound for the relative error of 10 3 ) or two decimals (which provides a bound for the absolute error of 10 2 ). Let x denote the true value and ˆx an approximation to x. Then the formulas are given by e x ˆx x (absolute error) r x e x x ˆx x x (relative error) Uncertainties of measurements are often described by expressions of the type x ˆx E x. This is meant to denote the estimate e x E x. Similarly, a statement that ˆx has d correct digits means that r x 10 d. Example 1.4. The following table illustrates the different notions. exact x π approximation ˆx absolute error e x relative error r x Example 1.5. Assume that we want to compute the sine of a value. Unfortunately, there is only a hand-held computer available which knows the four basic arithmetic operations, only. Therefore, we try to approximate the value of x sina for a given value of a. We know from analysis that Hence, we apply the approximation x sina a e x ˆx x max ã 0 5 a 3 6 a ˆx a 6 How large are the errors if we know that a o? It can be estimated that Similarly, the relative error can be estimated by So the error is much less than 0.1%! a 3 ã r x

5 1.2 Input Data and Computational Errors We consider a typical example. Assume that we want to compute the value of a scalar function for a given argument. We use the following notation: x true input data, f x true function value (provided by an analytic formula), ˆx approximate input data, ˆf approximate function evaluation (the algorithm used on the computer). Then the true result would be y f x. But what we really obtain is ŷ ˆf ˆx. So the absolute error amounts to e y ŷ y ˆf ˆx f x ˆf ˆx f ˆx computational error f ˆx f x error propagation The error propagation is a property of the problem at hand. In order to bound the error propagation the mathematical model must be constructed appropriately. The computational error is a property of the algorithm and the hardware (compiler, computer, computing environment, etc). While we do not have much influence on the latter, the correct design of the algorithm is the key to a successful solution of our problem. 1.3 Computational Errors We will observe two types of computational errors: truncation (or, discretization) errors and rounding errors. In order to analyze these errors we assume that even ˆx is given without errors, that is, we neglect the presence of errors in the input data. truncation error is the difference between the true result (for the actual input) and the result that would be produced by a given algorithm using exact arithmetic. It is due to approximations such as truncating an infinite series, replacing derivatives by finite differences, or terminating an iterative sequence before convergence. Rounding error is the difference between the result produced by a given algorithm using exact arithmetic and the result produced by the same algorithm using finite-precision, rounded arithmetic. Example 1.6. The derivative of a function f x at a point x is defined by f f x h f x x lim h 0 h 5

6 This infinite process must be approximated by a finite one. If h is sufficiently small, we expect the expression f a h f a D h f a h to be a good approximation of the derivative f a. We would even expect that, if h becomes smaller and smaller, this approximation becomes better and better. This intuitive behavior can be justified more rigorously. Assume that we write down the Taylor expansion of f at a: f a h f a h f a Inserting this expression in the definition of D h, we arrive at D h f a h 2 2 f a f a h O h 2 2 So we would expect that the error reduces proportional to h. Lets see what is happening on a real computer. Let us take the example f x e x and a 1. Then we have f 1 e The table below shows the results if we are using a calculator with eight digits of accuracy. h D h error The plot in Figure 1 shows the error curve. It is clearly seen that it consists of two parts. The above derive formula characterizes the truncation error. It is dominating as long as the step size h is not too small. Otherwise, the rounding error dominates. For completeness, both curves are added to the plot. The curve indicates that the rounding error behaves like O h 1. This is indeed expected. The difference f a h f a is computed up to a rounding error e in the order of magnitude of the machine accuracy. Therefore, the contribution of the rounding error can be estimated by e h O h Sensitivity and Conditioning Here we are interested in an estimation of the propagated error due to errors in the input data. This property is called sensitivity (if we think of the qualitative behavior). A quantification of 6

7 Computational error total truncation rounding error step size Figure 1: Error plot for the difference approximations in Example 1.6 the property is called conditioning (or, condition number) of a problem. The condition number is defined as the relation between the errors in the input data and those of the results: cond relative error of the result relative error in the input data The condition number is the factor by which errors in the input data are amplified because of the properties of the algorithm. A problem is called well-conditioned if the condition number is of moderate size. Otherwise, the problem is called ill-conditioned. Ill-conditioning is very often a severe problem in real life applications. If we consider our example of computing a function value of a scalar function, the condition number can be represented as follows: cond f ˆx f x f x ˆx x x y y x x x f x f x Here, we have used the estimation y x f x. In practice, the exact condition number is 7

8 usually not available. Instead, an estimation cond will be used such that we have the estimate relative error of the result cond relative error in the input data Example 1.7. Let us consider two simple examples: (i) f x x. A simple computation gives cond x 2 x x Consequently, the computation of the square root is very well conditioned. The relative error in the result is approximately halved. (ii) Consider the computation of f x tanx at x π 2. The derivative is given by f x 1 tan 2 x. This leads to a condition number cond x 1 tan 2 x tanx This condition number indicates that we will lose at least five digits. Since our input data has only five digits of accuracy, we expect that the result is completely unreliable. The following table shows that even the sign is not reliable: 1 2 ˆx tan ˆx Once the mathematical model is given, we do not any longer have any influence on its error propagation (or, with other words, its sensitivity). Let us now apply a numerical algorithm for solving it. Ideally, the final numerical result should not be much worse than the uncertainty which is already present in the input data. Therefore, an algorithm is called numerically stable if the computational errors are not much larger than the errors which are present because of the errors in the input data. Stability of an algorithm is, in some sense, analogous to the conditioning of a problem. We are interested in estimating the accuracy of a computed result. This means that we want to know how close the computed result comes to the true solution of the problem. From the discussion above, we find the following conclusions: Stability alone does not guarantee accuracy. Accuracy depends on conditioning of the problem as well as stability of the algorithm. Inaccuracy can result from applying a stable algorithm to an ill-conditioned problem or an unstable algorithm to a well-conditioned problem. Applying a stable algorithm to a well-conditioned problem yields an accurate solution. 8

9 1.5 A Few Words About Floating-Point Operations Computation on a computer means computation with finite precision. Almost all arithmetic operations lead to a rounding error. The representation of real numbers by floating point numbers leads to a common rounding error for all numbers as long as the numbers are not too large in magnitude (which is called overflow) or too small in absolute value (which is called underflow). The largest bound on the relative error is the accuracy of the computer. It is sometimes also called the machine epsilon ε mach. Today, the computer arithmetic is very much standardized by an ANSI-IEEE standard. The machine precision is (IEEE single precision), and (IEEE double precision), respectively. MATLAB uses internally IEEE double precision. But you should be aware of the fact that the numerical results can differ if you run your program on different hardware, even if all claim to use ANSI-IEEE arithmetic! According to its definition, the machine accuracy is the smallest rounding error in all arithmetic operations. But the more important question is how this errors propagate if very many (billions!!) of operations are carried out. For individual arithmetic operations, we have the following properties: Addition and subtraction ˆx ŷ x e x y e y x y e x e y The absolute errors in these operations will be added. This can lead to very large relative error if the final result x y is close to zero. This effect is called cancellation of leading digits. Consider as an example x , y Both values have a relative error of If we subtract these numbers, x y has a relative error of %! Multiplication ˆxŷ x 1 r x y 1 r y xy 1 r x 1 r y xy 1 r x r y Division Here, the relative errors are added. Hence, the multiplication is a very innocent operation with respect to error propagation. ˆx ŷ x y 1 r x 1 r y x y 1 r x r y The division has the same properties as the multiplication. Example 1.8. As an illustrative example of how to avoid catastrophic error propagation by cancellation, consider the solution of a quadratic equation ax 2 bx c 0. The well-known solution formula reads b b x 2 4ac 1 2 2a 9

10 Assume that there are two real roots. If b is positive, then the computation of the first root is subject to cancellation. If b is negative, the second root is subject to cancellation errors. The idea for constructing a better algorithm is to use Vieta s theorem: x 1 x 2 If b happens to be negative, compute x 1 by the former expression and then x 2 by x 2 c a c ax 1 If b is positive, the role of x 1 and x 2 will be exchanged. 1.6 Numerical Software High-quality mathematical software is available for solving most commonly occurring problems in scientific computing. The use of sophisticated, professionally written software has many advantages. We can concentrate ourselves on our problem. There will be no need to bother with highly difficult questions of numerical software architecture. Writing good software is an art. The following table contains an overview of sources for mathematical software. source description free? FMM From book by Forsyth/Malcolm/Moler yes HSL Harwell Subroutine Library partly IMSL Internat. Math. & Stat. Library no KMN From book Kahaner/Moler/Nash yes NAG Numerical Algorithms Group no Netlib Numerical library on the Internet yes NR From book Numerical Recipes no NUMAL Math. Centrum Amsterdam partly SLATEC From US Government Labs partly? SOL Systems Optimization Lab, Stanford no? TOMS ACM Trans. Math. Software yes It became more and more popular to use interactive environments for scientific computing. Such environments provide powerful mathematical capabilities, sophisticated graphics, and a high-level programming language for rapid prototyping. MATLAB is probably the most popular example. It is available for most personal computers, workstations, or even mainframes. There are also some free alternatives available. They include octave, RLaB, and Scilab. Sometimes symbolic computing environments are useful. They do not use numerical methods but abstract computer algebra tools. Prominent examples are Maple and Mathematica. There are also some free versions available. 10

Errors. Intensive Computation. Annalisa Massini 2017/2018

Errors. Intensive Computation. Annalisa Massini 2017/2018 Errors Intensive Computation Annalisa Massini 2017/2018 Intensive Computation - 2017/2018 2 References Scientific Computing: An Introductory Survey - Chapter 1 M.T. Heath http://heath.cs.illinois.edu/scicomp/notes/index.html

More information

Jim Lambers MAT 610 Summer Session Lecture 2 Notes

Jim Lambers MAT 610 Summer Session Lecture 2 Notes Jim Lambers MAT 610 Summer Session 2009-10 Lecture 2 Notes These notes correspond to Sections 2.2-2.4 in the text. Vector Norms Given vectors x and y of length one, which are simply scalars x and y, the

More information

Errors Intensive Computation

Errors Intensive Computation Errors Intensive Computation Annalisa Massini - 2015/2016 OVERVIEW Sources of Approimation Before computation modeling empirical measurements previous computations During computation truncation or discretization

More information

1 ERROR ANALYSIS IN COMPUTATION

1 ERROR ANALYSIS IN COMPUTATION 1 ERROR ANALYSIS IN COMPUTATION 1.2 Round-Off Errors & Computer Arithmetic (a) Computer Representation of Numbers Two types: integer mode (not used in MATLAB) floating-point mode x R ˆx F(β, t, l, u),

More information

Chapter 1 Error Analysis

Chapter 1 Error Analysis Chapter 1 Error Analysis Several sources of errors are important for numerical data processing: Experimental uncertainty: Input data from an experiment have a limited precision. Instead of the vector of

More information

Mathematical preliminaries and error analysis

Mathematical preliminaries and error analysis Mathematical preliminaries and error analysis Tsung-Ming Huang Department of Mathematics National Taiwan Normal University, Taiwan September 12, 2015 Outline 1 Round-off errors and computer arithmetic

More information

Notes on floating point number, numerical computations and pitfalls

Notes on floating point number, numerical computations and pitfalls Notes on floating point number, numerical computations and pitfalls November 6, 212 1 Floating point numbers An n-digit floating point number in base β has the form x = ±(.d 1 d 2 d n ) β β e where.d 1

More information

Lecture 7. Floating point arithmetic and stability

Lecture 7. Floating point arithmetic and stability Lecture 7 Floating point arithmetic and stability 2.5 Machine representation of numbers Scientific notation: 23 }{{} }{{} } 3.14159265 {{} }{{} 10 sign mantissa base exponent (significand) s m β e A floating

More information

1 Floating point arithmetic

1 Floating point arithmetic Introduction to Floating Point Arithmetic Floating point arithmetic Floating point representation (scientific notation) of numbers, for example, takes the following form.346 0 sign fraction base exponent

More information

Chapter 1 Mathematical Preliminaries and Error Analysis

Chapter 1 Mathematical Preliminaries and Error Analysis Numerical Analysis (Math 3313) 2019-2018 Chapter 1 Mathematical Preliminaries and Error Analysis Intended learning outcomes: Upon successful completion of this chapter, a student will be able to (1) list

More information

Chapter 1 Computer Arithmetic

Chapter 1 Computer Arithmetic Numerical Analysis (Math 9372) 2017-2016 Chapter 1 Computer Arithmetic 1.1 Introduction Numerical analysis is a way to solve mathematical problems by special procedures which use arithmetic operations

More information

Elements of Floating-point Arithmetic

Elements of Floating-point Arithmetic Elements of Floating-point Arithmetic Sanzheng Qiao Department of Computing and Software McMaster University July, 2012 Outline 1 Floating-point Numbers Representations IEEE Floating-point Standards Underflow

More information

Notes for Chapter 1 of. Scientific Computing with Case Studies

Notes for Chapter 1 of. Scientific Computing with Case Studies Notes for Chapter 1 of Scientific Computing with Case Studies Dianne P. O Leary SIAM Press, 2008 Mathematical modeling Computer arithmetic Errors 1999-2008 Dianne P. O'Leary 1 Arithmetic and Error What

More information

Floating Point Number Systems. Simon Fraser University Surrey Campus MACM 316 Spring 2005 Instructor: Ha Le

Floating Point Number Systems. Simon Fraser University Surrey Campus MACM 316 Spring 2005 Instructor: Ha Le Floating Point Number Systems Simon Fraser University Surrey Campus MACM 316 Spring 2005 Instructor: Ha Le 1 Overview Real number system Examples Absolute and relative errors Floating point numbers Roundoff

More information

Elements of Floating-point Arithmetic

Elements of Floating-point Arithmetic Elements of Floating-point Arithmetic Sanzheng Qiao Department of Computing and Software McMaster University July, 2012 Outline 1 Floating-point Numbers Representations IEEE Floating-point Standards Underflow

More information

Binary floating point

Binary floating point Binary floating point Notes for 2017-02-03 Why do we study conditioning of problems? One reason is that we may have input data contaminated by noise, resulting in a bad solution even if the intermediate

More information

1 Backward and Forward Error

1 Backward and Forward Error Math 515 Fall, 2008 Brief Notes on Conditioning, Stability and Finite Precision Arithmetic Most books on numerical analysis, numerical linear algebra, and matrix computations have a lot of material covering

More information

Arithmetic and Error. How does error arise? How does error arise? Notes for Part 1 of CMSC 460

Arithmetic and Error. How does error arise? How does error arise? Notes for Part 1 of CMSC 460 Notes for Part 1 of CMSC 460 Dianne P. O Leary Preliminaries: Mathematical modeling Computer arithmetic Errors 1999-2006 Dianne P. O'Leary 1 Arithmetic and Error What we need to know about error: -- how

More information

An Introduction to Numerical Analysis

An Introduction to Numerical Analysis An Introduction to Numerical Analysis Department of Mathematical Sciences, NTNU 21st august 2012 Practical issues webpage: http://wiki.math.ntnu.no/tma4215/2012h/start Lecturer:, elenac@math.ntnu.no Lecures:

More information

Numerical Algorithms. IE 496 Lecture 20

Numerical Algorithms. IE 496 Lecture 20 Numerical Algorithms IE 496 Lecture 20 Reading for This Lecture Primary Miller and Boxer, Pages 124-128 Forsythe and Mohler, Sections 1 and 2 Numerical Algorithms Numerical Analysis So far, we have looked

More information

Applied Mathematics 205. Unit 0: Overview of Scientific Computing. Lecturer: Dr. David Knezevic

Applied Mathematics 205. Unit 0: Overview of Scientific Computing. Lecturer: Dr. David Knezevic Applied Mathematics 205 Unit 0: Overview of Scientific Computing Lecturer: Dr. David Knezevic Scientific Computing Computation is now recognized as the third pillar of science (along with theory and experiment)

More information

Queens College, CUNY, Department of Computer Science Numerical Methods CSCI 361 / 761 Spring 2018 Instructor: Dr. Sateesh Mane.

Queens College, CUNY, Department of Computer Science Numerical Methods CSCI 361 / 761 Spring 2018 Instructor: Dr. Sateesh Mane. Queens College, CUNY, Department of Computer Science Numerical Methods CSCI 361 / 761 Spring 2018 Instructor: Dr. Sateesh Mane c Sateesh R. Mane 2018 3 Lecture 3 3.1 General remarks March 4, 2018 This

More information

Math 411 Preliminaries

Math 411 Preliminaries Math 411 Preliminaries Provide a list of preliminary vocabulary and concepts Preliminary Basic Netwon s method, Taylor series expansion (for single and multiple variables), Eigenvalue, Eigenvector, Vector

More information

Lies My Calculator and Computer Told Me

Lies My Calculator and Computer Told Me Lies My Calculator and Computer Told Me 2 LIES MY CALCULATOR AND COMPUTER TOLD ME Lies My Calculator and Computer Told Me See Section.4 for a discussion of graphing calculators and computers with graphing

More information

Tu: 9/3/13 Math 471, Fall 2013, Section 001 Lecture 1

Tu: 9/3/13 Math 471, Fall 2013, Section 001 Lecture 1 Tu: 9/3/13 Math 71, Fall 2013, Section 001 Lecture 1 1 Course intro Notes : Take attendance. Instructor introduction. Handout : Course description. Note the exam days (and don t be absent). Bookmark the

More information

Chapter 1: Introduction and mathematical preliminaries

Chapter 1: Introduction and mathematical preliminaries Chapter 1: Introduction and mathematical preliminaries Evy Kersalé September 26, 2011 Motivation Most of the mathematical problems you have encountered so far can be solved analytically. However, in real-life,

More information

Numerical Methods - Lecture 2. Numerical Methods. Lecture 2. Analysis of errors in numerical methods

Numerical Methods - Lecture 2. Numerical Methods. Lecture 2. Analysis of errors in numerical methods Numerical Methods - Lecture 1 Numerical Methods Lecture. Analysis o errors in numerical methods Numerical Methods - Lecture Why represent numbers in loating point ormat? Eample 1. How a number 56.78 can

More information

Introduction to Finite Di erence Methods

Introduction to Finite Di erence Methods Introduction to Finite Di erence Methods ME 448/548 Notes Gerald Recktenwald Portland State University Department of Mechanical Engineering gerry@pdx.edu ME 448/548: Introduction to Finite Di erence Approximation

More information

What Every Programmer Should Know About Floating-Point Arithmetic DRAFT. Last updated: November 3, Abstract

What Every Programmer Should Know About Floating-Point Arithmetic DRAFT. Last updated: November 3, Abstract What Every Programmer Should Know About Floating-Point Arithmetic Last updated: November 3, 2014 Abstract The article provides simple answers to the common recurring questions of novice programmers about

More information

MAT128A: Numerical Analysis Lecture Three: Condition Numbers

MAT128A: Numerical Analysis Lecture Three: Condition Numbers MAT128A: Numerical Analysis Lecture Three: Condition Numbers October 1, 2018 Lecture 1 October 1, 2018 1 / 26 An auspicious example Last time, we saw that the naive evaluation of the function f (x) = 1

More information

Lecture Notes 7, Math/Comp 128, Math 250

Lecture Notes 7, Math/Comp 128, Math 250 Lecture Notes 7, Math/Comp 128, Math 250 Misha Kilmer Tufts University October 23, 2005 Floating Point Arithmetic We talked last time about how the computer represents floating point numbers. In a floating

More information

ECS 231 Computer Arithmetic 1 / 27

ECS 231 Computer Arithmetic 1 / 27 ECS 231 Computer Arithmetic 1 / 27 Outline 1 Floating-point numbers and representations 2 Floating-point arithmetic 3 Floating-point error analysis 4 Further reading 2 / 27 Outline 1 Floating-point numbers

More information

Introduction to Scientific Computing

Introduction to Scientific Computing (Lecture 2: Machine precision and condition number) B. Rosić, T.Moshagen Institute of Scientific Computing General information :) 13 homeworks (HW) Work in groups of 2 or 3 people Each HW brings maximally

More information

Lecture 28 The Main Sources of Error

Lecture 28 The Main Sources of Error Lecture 28 The Main Sources of Error Truncation Error Truncation error is defined as the error caused directly by an approximation method For instance, all numerical integration methods are approximations

More information

Introduction to Scientific Computing Languages

Introduction to Scientific Computing Languages 1 / 19 Introduction to Scientific Computing Languages Prof. Paolo Bientinesi pauldj@aices.rwth-aachen.de Numerical Representation 2 / 19 Numbers 123 = (first 40 digits) 29 4.241379310344827586206896551724137931034...

More information

MATH ASSIGNMENT 03 SOLUTIONS

MATH ASSIGNMENT 03 SOLUTIONS MATH444.0 ASSIGNMENT 03 SOLUTIONS 4.3 Newton s method can be used to compute reciprocals, without division. To compute /R, let fx) = x R so that fx) = 0 when x = /R. Write down the Newton iteration for

More information

FLOATING POINT ARITHMETHIC - ERROR ANALYSIS

FLOATING POINT ARITHMETHIC - ERROR ANALYSIS FLOATING POINT ARITHMETHIC - ERROR ANALYSIS Brief review of floating point arithmetic Model of floating point arithmetic Notation, backward and forward errors Roundoff errors and floating-point arithmetic

More information

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 5. Ax = b.

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 5. Ax = b. CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 5 GENE H GOLUB Suppose we want to solve We actually have an approximation ξ such that 1 Perturbation Theory Ax = b x = ξ + e The question is, how

More information

8.5 Taylor Polynomials and Taylor Series

8.5 Taylor Polynomials and Taylor Series 8.5. TAYLOR POLYNOMIALS AND TAYLOR SERIES 50 8.5 Taylor Polynomials and Taylor Series Motivating Questions In this section, we strive to understand the ideas generated by the following important questions:

More information

A Few Concepts from Numerical Analysis

A Few Concepts from Numerical Analysis 2 A Few Concepts from Numerical Analysis A systematic treatment of numerical methods is provided in conventional courses and textbooks on numerical analysis. But a few very common issues, that emerge in

More information

FLOATING POINT ARITHMETHIC - ERROR ANALYSIS

FLOATING POINT ARITHMETHIC - ERROR ANALYSIS FLOATING POINT ARITHMETHIC - ERROR ANALYSIS Brief review of floating point arithmetic Model of floating point arithmetic Notation, backward and forward errors 3-1 Roundoff errors and floating-point arithmetic

More information

INTRODUCTION TO COMPUTATIONAL MATHEMATICS

INTRODUCTION TO COMPUTATIONAL MATHEMATICS INTRODUCTION TO COMPUTATIONAL MATHEMATICS Course Notes for CM 271 / AMATH 341 / CS 371 Fall 2007 Instructor: Prof. Justin Wan School of Computer Science University of Waterloo Course notes by Prof. Hans

More information

Numerical Analysis. Yutian LI. 2018/19 Term 1 CUHKSZ. Yutian LI (CUHKSZ) Numerical Analysis 2018/19 1 / 41

Numerical Analysis. Yutian LI. 2018/19 Term 1 CUHKSZ. Yutian LI (CUHKSZ) Numerical Analysis 2018/19 1 / 41 Numerical Analysis Yutian LI CUHKSZ 2018/19 Term 1 Yutian LI (CUHKSZ) Numerical Analysis 2018/19 1 / 41 Reference Books BF R. L. Burden and J. D. Faires, Numerical Analysis, 9th edition, Thomsom Brooks/Cole,

More information

Introduction to Scientific Computing Languages

Introduction to Scientific Computing Languages 1 / 21 Introduction to Scientific Computing Languages Prof. Paolo Bientinesi pauldj@aices.rwth-aachen.de Numerical Representation 2 / 21 Numbers 123 = (first 40 digits) 29 4.241379310344827586206896551724137931034...

More information

THE SECANT METHOD. q(x) = a 0 + a 1 x. with

THE SECANT METHOD. q(x) = a 0 + a 1 x. with THE SECANT METHOD Newton s method was based on using the line tangent to the curve of y = f (x), with the point of tangency (x 0, f (x 0 )). When x 0 α, the graph of the tangent line is approximately the

More information

Numerical Methods. King Saud University

Numerical Methods. King Saud University Numerical Methods King Saud University Aims In this lecture, we will... Introduce the topic of numerical methods Consider the Error analysis and sources of errors Introduction A numerical method which

More information

Numerical Analysis and Computing

Numerical Analysis and Computing Numerical Analysis and Computing Lecture Notes #02 Calculus Review; Computer Artihmetic and Finite Precision; and Convergence; Joe Mahaffy, mahaffy@math.sdsu.edu Department of Mathematics Dynamical Systems

More information

MyMathLab for School Precalculus Graphical, Numerical, Algebraic Common Core Edition 2016

MyMathLab for School Precalculus Graphical, Numerical, Algebraic Common Core Edition 2016 A Correlation of MyMathLab for School Precalculus Common Core Edition 2016 to the Tennessee Mathematics Standards Approved July 30, 2010 Bid Category 13-090-10 , Standard 1 Mathematical Processes Course

More information

NUMERICAL MATHEMATICS & COMPUTING 6th Edition

NUMERICAL MATHEMATICS & COMPUTING 6th Edition NUMERICAL MATHEMATICS & COMPUTING 6th Edition Ward Cheney/David Kincaid c UT Austin Engage Learning: Thomson-Brooks/Cole www.engage.com www.ma.utexas.edu/cna/nmc6 September 1, 2011 2011 1 / 42 1.1 Mathematical

More information

NUMERICAL AND STATISTICAL COMPUTING (MCA-202-CR)

NUMERICAL AND STATISTICAL COMPUTING (MCA-202-CR) NUMERICAL AND STATISTICAL COMPUTING (MCA-202-CR) Autumn Session UNIT 1 Numerical analysis is the study of algorithms that uses, creates and implements algorithms for obtaining numerical solutions to problems

More information

QUADRATIC PROGRAMMING?

QUADRATIC PROGRAMMING? QUADRATIC PROGRAMMING? WILLIAM Y. SIT Department of Mathematics, The City College of The City University of New York, New York, NY 10031, USA E-mail: wyscc@cunyvm.cuny.edu This is a talk on how to program

More information

Introduction, basic but important concepts

Introduction, basic but important concepts Introduction, basic but important concepts Felix Kubler 1 1 DBF, University of Zurich and Swiss Finance Institute October 7, 2017 Felix Kubler Comp.Econ. Gerzensee, Ch1 October 7, 2017 1 / 31 Economics

More information

MATH2071: LAB #5: Norms, Errors and Condition Numbers

MATH2071: LAB #5: Norms, Errors and Condition Numbers MATH2071: LAB #5: Norms, Errors and Condition Numbers 1 Introduction Introduction Exercise 1 Vector Norms Exercise 2 Matrix Norms Exercise 3 Compatible Matrix Norms Exercise 4 More on the Spectral Radius

More information

TEACHING NUMERICAL LINEAR ALGEBRA AT THE UNDERGRADUATE LEVEL by Biswa Nath Datta Department of Mathematical Sciences Northern Illinois University

TEACHING NUMERICAL LINEAR ALGEBRA AT THE UNDERGRADUATE LEVEL by Biswa Nath Datta Department of Mathematical Sciences Northern Illinois University TEACHING NUMERICAL LINEAR ALGEBRA AT THE UNDERGRADUATE LEVEL by Biswa Nath Datta Department of Mathematical Sciences Northern Illinois University DeKalb, IL 60115 E-mail: dattab@math.niu.edu What is Numerical

More information

HIGH SCHOOL MATH CURRICULUM GRADE ELEVEN ALGEBRA 2 & TRIGONOMETRY N

HIGH SCHOOL MATH CURRICULUM GRADE ELEVEN ALGEBRA 2 & TRIGONOMETRY N VALLEY CENTRAL SCHOOL DISTRICT 944 STATE ROUTE 17K MONTGOMERY, NY 12549 Telephone Number: (845) 457-2400 ext. 8121 FAX NUMBER: (845) 457-4254 HIGH SCHOOL MATH CURRICULUM GRADE ELEVEN ALGEBRA 2 & TRIGONOMETRY

More information

ALGEBRA GRADE 7 MINNESOTA ACADEMIC STANDARDS CORRELATED TO MOVING WITH MATH. Part B Student Book Skill Builders (SB)

ALGEBRA GRADE 7 MINNESOTA ACADEMIC STANDARDS CORRELATED TO MOVING WITH MATH. Part B Student Book Skill Builders (SB) MINNESOTA ACADEMIC STANDARDS CORRELATED TO MOVING WITH MATH ALGEBRA GRADE 7 NUMBER AND OPERATION Read, write, represent and compare positive and negative rational numbers, expressed as integers, fractions

More information

Numerical Methods in Physics and Astrophysics

Numerical Methods in Physics and Astrophysics Kostas Kokkotas 2 November 6, 2007 2 kostas.kokkotas@uni-tuebingen.de http://www.tat.physik.uni-tuebingen.de/kokkotas Kostas Kokkotas 3 Error Analysis Definition : Suppose that x is an approximation to

More information

Floating-point Computation

Floating-point Computation Chapter 2 Floating-point Computation 21 Positional Number System An integer N in a number system of base (or radix) β may be written as N = a n β n + a n 1 β n 1 + + a 1 β + a 0 = P n (β) where a i are

More information

EAD 115. Numerical Solution of Engineering and Scientific Problems. David M. Rocke Department of Applied Science

EAD 115. Numerical Solution of Engineering and Scientific Problems. David M. Rocke Department of Applied Science EAD 115 Numerical Solution of Engineering and Scientific Problems David M. Rocke Department of Applied Science Computer Representation of Numbers Counting numbers (unsigned integers) are the numbers 0,

More information

Math 471. Numerical methods Introduction

Math 471. Numerical methods Introduction Math 471. Numerical methods Introduction Section 1.1 1.4 of Bradie 1.1 Algorithms Here is an analogy between Numerical Methods and Gastronomy: Calculus, Lin Alg., Diff. eq. Ingredients Algorithm Recipe

More information

Today s lecture. WEST VIRGINIA UNIVERSITY Physics

Today s lecture. WEST VIRGINIA UNIVERSITY Physics Today s lecture Units, Estimations, Graphs, Trigonometry: Units - Standards of Length, Mass, and Time Dimensional Analysis Uncertainty and significant digits Order of magnitude estimations Coordinate Systems

More information

Remainders. We learned how to multiply and divide in elementary

Remainders. We learned how to multiply and divide in elementary Remainders We learned how to multiply and divide in elementary school. As adults we perform division mostly by pressing the key on a calculator. This key supplies the quotient. In numerical analysis and

More information

Can ANSYS Mechanical Handle My Required Modeling Precision?

Can ANSYS Mechanical Handle My Required Modeling Precision? Can ANSYS Mechanical Handle My Required Modeling Precision? 2/8/2017 www.telescope-optics.net Recently, PADT received the following inquiry from a scientist interested in using ANSYS for telescope application

More information

Lecture for Week 2 (Secs. 1.3 and ) Functions and Limits

Lecture for Week 2 (Secs. 1.3 and ) Functions and Limits Lecture for Week 2 (Secs. 1.3 and 2.2 2.3) Functions and Limits 1 First let s review what a function is. (See Sec. 1 of Review and Preview.) The best way to think of a function is as an imaginary machine,

More information

2 Discrete Dynamical Systems (DDS)

2 Discrete Dynamical Systems (DDS) 2 Discrete Dynamical Systems (DDS) 2.1 Basics A Discrete Dynamical System (DDS) models the change (or dynamics) of single or multiple populations or quantities in which the change occurs deterministically

More information

CHAPTER 3. Iterative Methods

CHAPTER 3. Iterative Methods CHAPTER 3 Iterative Methods As we have seen in the previous two chapters, even for problems, which are theoretically well understood, such as computing the square root, one cannot provide the perfect answer

More information

Algebra Year 10. Language

Algebra Year 10. Language Algebra Year 10 Introduction In Algebra we do Maths with numbers, but some of those numbers are not known. They are represented with letters, and called unknowns, variables or, most formally, literals.

More information

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 6

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 6 CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 6 GENE H GOLUB Issues with Floating-point Arithmetic We conclude our discussion of floating-point arithmetic by highlighting two issues that frequently

More information

6 Linear Systems of Equations

6 Linear Systems of Equations 6 Linear Systems of Equations Read sections 2.1 2.3, 2.4.1 2.4.5, 2.4.7, 2.7 Review questions 2.1 2.37, 2.43 2.67 6.1 Introduction When numerically solving two-point boundary value problems, the differential

More information

1.1 COMPUTER REPRESENTATION OF NUM- BERS, REPRESENTATION ERRORS

1.1 COMPUTER REPRESENTATION OF NUM- BERS, REPRESENTATION ERRORS Chapter 1 NUMBER REPRESENTATION, ERROR ANALYSIS 1.1 COMPUTER REPRESENTATION OF NUM- BERS, REPRESENTATION ERRORS Floating-point representation x t,r of a number x: x t,r = m t P cr, where: P - base (the

More information

Chapter 1 Mathematical Preliminaries and Error Analysis

Chapter 1 Mathematical Preliminaries and Error Analysis Chapter 1 Mathematical Preliminaries and Error Analysis Per-Olof Persson persson@berkeley.edu Department of Mathematics University of California, Berkeley Math 128A Numerical Analysis Limits and Continuity

More information

2. Introduction to commutative rings (continued)

2. Introduction to commutative rings (continued) 2. Introduction to commutative rings (continued) 2.1. New examples of commutative rings. Recall that in the first lecture we defined the notions of commutative rings and field and gave some examples of

More information

This ensures that we walk downhill. For fixed λ not even this may be the case.

This ensures that we walk downhill. For fixed λ not even this may be the case. Gradient Descent Objective Function Some differentiable function f : R n R. Gradient Descent Start with some x 0, i = 0 and learning rate λ repeat x i+1 = x i λ f(x i ) until f(x i+1 ) ɛ Line Search Variant

More information

One-Sided Difference Formula for the First Derivative

One-Sided Difference Formula for the First Derivative POLYTECHNIC UNIVERSITY Department of Computer and Information Science One-Sided Difference Formula for the First Derivative K. Ming Leung Abstract: Derive a one-sided formula for the first derive of a

More information

Graphical Analysis and Errors MBL

Graphical Analysis and Errors MBL Graphical Analysis and Errors MBL I Graphical Analysis Graphs are vital tools for analyzing and displaying data Graphs allow us to explore the relationship between two quantities -- an independent variable

More information

Essentials of Intermediate Algebra

Essentials of Intermediate Algebra Essentials of Intermediate Algebra BY Tom K. Kim, Ph.D. Peninsula College, WA Randy Anderson, M.S. Peninsula College, WA 9/24/2012 Contents 1 Review 1 2 Rules of Exponents 2 2.1 Multiplying Two Exponentials

More information

Introduction and mathematical preliminaries

Introduction and mathematical preliminaries Chapter Introduction and mathematical preliminaries Contents. Motivation..................................2 Finite-digit arithmetic.......................... 2.3 Errors in numerical calculations.....................

More information

MAT 460: Numerical Analysis I. James V. Lambers

MAT 460: Numerical Analysis I. James V. Lambers MAT 460: Numerical Analysis I James V. Lambers January 31, 2013 2 Contents 1 Mathematical Preliminaries and Error Analysis 7 1.1 Introduction............................ 7 1.1.1 Error Analysis......................

More information

CHAPTER 10 Zeros of Functions

CHAPTER 10 Zeros of Functions CHAPTER 10 Zeros of Functions An important part of the maths syllabus in secondary school is equation solving. This is important for the simple reason that equations are important a wide range of problems

More information

Assume that you have made n different measurements of a quantity x. Usually the results of these measurements will vary; call them x 1

Assume that you have made n different measurements of a quantity x. Usually the results of these measurements will vary; call them x 1 #1 $ http://www.physics.fsu.edu/users/ng/courses/phy2048c/lab/appendixi/app1.htm Appendix I: Estimates for the Reliability of Measurements In any measurement there is always some error or uncertainty in

More information

Deep Learning. Authors: I. Goodfellow, Y. Bengio, A. Courville. Chapter 4: Numerical Computation. Lecture slides edited by C. Yim. C.

Deep Learning. Authors: I. Goodfellow, Y. Bengio, A. Courville. Chapter 4: Numerical Computation. Lecture slides edited by C. Yim. C. Chapter 4: Numerical Computation Deep Learning Authors: I. Goodfellow, Y. Bengio, A. Courville Lecture slides edited by 1 Chapter 4: Numerical Computation 4.1 Overflow and Underflow 4.2 Poor Conditioning

More information

Hani Mehrpouyan, California State University, Bakersfield. Signals and Systems

Hani Mehrpouyan, California State University, Bakersfield. Signals and Systems Hani Mehrpouyan, Department of Electrical and Computer Engineering, California State University, Bakersfield Lecture 1 (Intro, History and Background) April 4 th, 2013 The material in these lectures is

More information

Accurate polynomial evaluation in floating point arithmetic

Accurate polynomial evaluation in floating point arithmetic in floating point arithmetic Université de Perpignan Via Domitia Laboratoire LP2A Équipe de recherche en Informatique DALI MIMS Seminar, February, 10 th 2006 General motivation Provide numerical algorithms

More information

Numerical Linear Algebra

Numerical Linear Algebra Schedule Prerequisite Preliminaries Errors and Algorithms Numerical Linear Algebra Kim, Hyun-Min Department of Mathematics, Pusan National University E-mail:hyunmin@pusan.ac.kr Phone: 510-1060, 2596, 010-3833-8200

More information

1. Introduction to commutative rings and fields

1. Introduction to commutative rings and fields 1. Introduction to commutative rings and fields Very informally speaking, a commutative ring is a set in which we can add, subtract and multiply elements so that the usual laws hold. A field is a commutative

More information

CS760, S. Qiao Part 3 Page 1. Kernighan and Pike [5] present the following general guidelines for testing software.

CS760, S. Qiao Part 3 Page 1. Kernighan and Pike [5] present the following general guidelines for testing software. CS760, S. Qiao Part 3 Page 1 Testing 1 General Guidelines Kernighan and Pike [5] present the following general guidelines for testing software. Know exactly what you are testing and what results you expect.

More information

1.1.1 Algebraic Operations

1.1.1 Algebraic Operations 1.1.1 Algebraic Operations We need to learn how our basic algebraic operations interact. When confronted with many operations, we follow the order of operations: Parentheses Exponentials Multiplication

More information

Measurement and Uncertainty

Measurement and Uncertainty Measurement and Uncertainty Name: Date: Block: There is uncertainty in every measurement due to of accuracy and precision. Accuracy: how close the instrument measures to an accepted. Precision: how closely

More information

MATHEMATICS. Perform a series of transformations and/or dilations to a figure. A FAMILY GUIDE FOR STUDENT SUCCESS 17

MATHEMATICS. Perform a series of transformations and/or dilations to a figure. A FAMILY GUIDE FOR STUDENT SUCCESS 17 MATHEMATICS In grade 8, your child will focus on three critical areas. The first is formulating and reasoning about expressions and equations, including modeling an association in bivariate data with a

More information

Mathematics for Engineers. Numerical mathematics

Mathematics for Engineers. Numerical mathematics Mathematics for Engineers Numerical mathematics Integers Determine the largest representable integer with the intmax command. intmax ans = int32 2147483647 2147483647+1 ans = 2.1475e+09 Remark The set

More information

Numerical Computation for Deep Learning

Numerical Computation for Deep Learning Numerical Computation for Deep Learning Lecture slides for Chapter 4 of Deep Learning www.deeplearningbook.org Ian Goodfellow Last modified 2017-10-14 Thanks to Justin Gilmer and Jacob Buckman for helpful

More information

Bindel, Fall 2011 Intro to Scientific Computing (CS 3220) Week 3: Wednesday, Jan 9

Bindel, Fall 2011 Intro to Scientific Computing (CS 3220) Week 3: Wednesday, Jan 9 Problem du jour Week 3: Wednesday, Jan 9 1. As a function of matrix dimension, what is the asymptotic complexity of computing a determinant using the Laplace expansion (cofactor expansion) that you probably

More information

Linear algebra for MATH2601 Numerical methods

Linear algebra for MATH2601 Numerical methods Linear algebra for MATH2601 Numerical methods László Erdős August 12, 2000 Contents 1 Introduction 3 1.1 Typesoferrors... 4 1.1.1 Rounding errors... 5 1.1.2 Truncationerrors... 6 1.1.3 Conditioningerrors...

More information

JUST THE MATHS UNIT NUMBER 1.6. ALGEBRA 6 (Formulae and algebraic equations) A.J.Hobson

JUST THE MATHS UNIT NUMBER 1.6. ALGEBRA 6 (Formulae and algebraic equations) A.J.Hobson JUST THE MATHS UNIT NUMBER 1.6 ALGEBRA 6 (Formulae and algebraic equations) by A.J.Hobson 1.6.1 Transposition of formulae 1.6. of linear equations 1.6.3 of quadratic equations 1.6. Exercises 1.6.5 Answers

More information

PART I Lecture Notes on Numerical Solution of Root Finding Problems MATH 435

PART I Lecture Notes on Numerical Solution of Root Finding Problems MATH 435 PART I Lecture Notes on Numerical Solution of Root Finding Problems MATH 435 Professor Biswa Nath Datta Department of Mathematical Sciences Northern Illinois University DeKalb, IL. 60115 USA E mail: dattab@math.niu.edu

More information

Next topics: Solving systems of linear equations

Next topics: Solving systems of linear equations Next topics: Solving systems of linear equations 1 Gaussian elimination (today) 2 Gaussian elimination with partial pivoting (Week 9) 3 The method of LU-decomposition (Week 10) 4 Iterative techniques:

More information

Review of matrices. Let m, n IN. A rectangle of numbers written like A =

Review of matrices. Let m, n IN. A rectangle of numbers written like A = Review of matrices Let m, n IN. A rectangle of numbers written like a 11 a 12... a 1n a 21 a 22... a 2n A =...... a m1 a m2... a mn where each a ij IR is called a matrix with m rows and n columns or an

More information

ES.1803 Topic 13 Notes Jeremy Orloff

ES.1803 Topic 13 Notes Jeremy Orloff ES.1803 Topic 13 Notes Jeremy Orloff 13 Vector Spaces, matrices and linearity 13.1 Goals 1. Know the definition of a vector space and how to show that a given set is a vector space. 2. Know the meaning

More information

Chapter 1A -- Real Numbers. iff. Math Symbols: Sets of Numbers

Chapter 1A -- Real Numbers. iff. Math Symbols: Sets of Numbers Fry Texas A&M University! Fall 2016! Math 150 Notes! Section 1A! Page 1 Chapter 1A -- Real Numbers Math Symbols: iff or Example: Let A = {2, 4, 6, 8, 10, 12, 14, 16,...} and let B = {3, 6, 9, 12, 15, 18,

More information