Scribd
Upload
10 views

MATH685_Sp10_lec1

The document provides an introduction to scientific computing, covering essential concepts such as numerical methods, sources of errors, and the importance of stability and accuracy in algorithms. It discusses the well-posedness of problems, conditioning, and the impact of floating-point representation on computations. Additionally, it includes resources for MATLAB and OCTAVE, as well as techniques for error analysis and algorithm implementation.

Uploaded by

nguyenxuandung457
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
10 views

MATH685_Sp10_lec1

The document provides an introduction to scientific computing, covering essential concepts such as numerical methods, sources of errors, and the importance of stability and accuracy in algorithms. It discusses the well-posedness of problems, conditioning, and the impact of floating-point representation on computations. Additionally, it includes resources for MATLAB and OCTAVE, as well as techniques for error analysis and algorithm implementation.

Uploaded by

nguyenxuandung457
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 26

MATH 685/CSI 700

Lecture Notes
Lecture 1.
Intro to Scientific Computing
Useful info
 Course website:
http://math.gmu.edu/~memelian/teaching/Spring10
 MATLAB instructions:
http://math.gmu.edu/introtomatlab.htm
 Mathworks, the creator of MATLAB: http://www.mathworks.com
 OCTAVE = free MATLAB clone
Available for download at http://octave.sourceforge.net/
Phương pháp tính
 Thiết kế và phân tích các thuật toán số giải các bài toán trong khoa
học và kỹ thuật
 Thực hiện việc chuyển đổi đại lượng liên tục sang đại lượng rời rạc
(chẳng hạn như khoa học máy tính)
 Xem xét sự ảnh hưởng của phép tính gần đúng và thực hiện phân
tích sai số
 Có mặt khắp nơi trong các bài toán mô phỏng và thuật toán hiện đại
mô hình hóa các hiện tượng tự nhiên và trong các ứng dụng kỹ
thuật
 Liên quan chặt chẽ đến Giải tích số
Computational problems:
attack strategy
 Phát triển mô hình toán học (thường đòi hỏi sự kết hợp giữa kỹ
năng toán học và một số kiến thức tiên nghiệm về hệ thống)
 Đưa ra thuật toán số (kỹ năng phân tích số)
 Triển khai thuật toán (kỹ năng phần mềm)
 Run, debug, test the software

 Visualize the results Mathematical modeling

 Interpret and validate the results


Computational problems:
well-posedness
 The problem is well-posed, if
(a) solution exists
(b) it is unique
(c) it depends continuously on problem data
The problem can be well-posed, but still sensitive to perturbations.
The algorithm should attempt to simplify the problem, but not make
sensitivity worse than it already is.
e
at
Simplification strategies: xim e !
p ro n b ay
Infinite finite ap ca w
ly io n this
Nonlinear linear On olut ed
s ain
High-order low-order t
ob
Sources of numerical errors
 Before computation
 modeling approximations
 empirical measurements, human errors Cannot be controlled
 previous computations
 During computation
 truncation or discretization Can be controlled through
 Rounding errors error analysis
 Accuracy depends on both, but we can only control the second part
 Uncertainty in input may be amplified by problem
 Perturbations during computation may be amplified by algorithm

Abs_error = approx_value – true_value


Rel_error = abs_error/true_value
Approx_value = (true_value)x(1+rel_error)
Sources of numerical errors
 Propagated vs. computational error
 x = exact value, y = approx. value
 F = exact function, G = its approximation
 G(y) – F(x) = [G(y) - F(y)] + [F(y) - F(x)]

Total error = Computational error: Propagated data error:


+
affected by algorithm not affected by algorithm
 Rounding vs. truncation error
 Rounding error: introduced by finite precision calculations in the
computer arithmetic
 Truncation error: introduced by algorithm via problem simplification, e.g.
series truncation, iterative process truncation etc.

Computational error = Truncation error + rounding error


Backward vs. forward errors
Backward error analysis
 How much must original problem change to give result actually
obtained?

 How much data error in input would explain all error in


computed result?

 Approximate solution is good if it is the exact solution to a nearby


problem

 Backward error is often easier to estimate than forward error


Backward vs. forward errors
Conditioning
 Well-conditioned (insensitive) problem: small relative change in
input gives commensurate relative change in the solution
 Ill-conditioned (sensitive): relative change in output is much larger
than that in the input data
 Condition number = measure of sensitivity

 Condition number = |rel. forward error| / |rel. backward error|


= amplification factor
Conditioning
Stability
 Algorithm is stable if result produced is relatively insensitive to
perturbations during computation

 Stability of algorithms is analogous to conditioning of


problems

 From point of view of backward error analysis, algorithm is stable


if result produced is exact solution to nearby problem

 For stable algorithm, effect of computational error is no worse


than effect of small data error in input
Accuracy
 Accuracy : closeness of computed solution to true solution
of problem

 Stability alone does not guarantee accurate results

 Accuracy depends on conditioning of problem as well as


stability of algorithm

 Inaccuracy can result from applying stable algorithm to


ill-conditioned problem or unstable algorithm to well-conditioned
problem

 Applying stable algorithm to well-conditioned problem


yields accurate solution
Floating point representation
Floating point systems
Normalized representation

Not all numbers can be represented this way, those that can are called
machine numbers
Rounding rules
 If real number x is not exactly representable, then it is approximated by
“nearby” floating-point number fl(x)

 This process is called rounding, and error introduced is called rounding


error

 Two commonly used rounding rules


 chop: truncate base- expansion of x after (p − 1)st digit;
also called round toward zero
 round to nearest : fl(x) is nearest floating-point number to x, using floating-
point number whose last stored digit is even in case of tie; also called round
to even

 Round to nearest is most accurate, and is default rounding rule in IEEE


systems
Floating point arithmetic
Machine precision
Floating point operations
Summing series in floating-
point arithmetic
Loss of significance
Loss of significance
Loss of significance
Loss of significance

You might also like