Scribd
Upload
144 views2 pages

DHB Java

This document reviews the book "Java Number Cruncher" by Ronald Mak, which teaches Java programmers the basics of numerical computing. It summarizes that Java is now a viable option for numerical computing thanks to just-in-time compilers. The book covers basic numerical methods like floating point arithmetic, root-finding, interpolation, integration, and solving differential equations. It focuses on concepts over implementation details. Later chapters discuss matrix computations, high-precision arithmetic, and applications like factoring and computing constants/functions to high precision. The review concludes that the book would make an excellent introductory course to numerical computing for computer scientists, engineers and scientists.

Uploaded by

bernasek
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
144 views2 pages

DHB Java

This document reviews the book "Java Number Cruncher" by Ronald Mak, which teaches Java programmers the basics of numerical computing. It summarizes that Java is now a viable option for numerical computing thanks to just-in-time compilers. The book covers basic numerical methods like floating point arithmetic, root-finding, interpolation, integration, and solving differential equations. It focuses on concepts over implementation details. Later chapters discuss matrix computations, high-precision arithmetic, and applications like factoring and computing constants/functions to high precision. The review concludes that the book would make an excellent introductory course to numerical computing for computer scientists, engineers and scientists.

Uploaded by

bernasek
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Java Meets Numerical Analysis Review of Java Number Cruncher: The Java Programmers Guide to Numerical Computing by Ronald

Mak Review by David H. Bailey, Lawrence Berkeley National Laboratory To appear in Scientic Programming

Introduction Java for numerical computing? Are you kidding!? Until recently, such a combination would indeed draw well-deserved howls of laughter from serious practitioners of numerical computing. The huge overhead of Javas traditional interpreted-code implementation has resulted in run times that have been typically at least an order of magnitude higher than traditional languages such as Fortran-90 or C/C++. As a result, Java has been a non-starter for large, numerically intensive scientic computation. But times are changing. Partly because of the extensive interest in Java programming in the business and Internet world, just-in-time compilers and the like have been developed that result in much faster execution times, nearly competitive with C and Fortran code in many cases. As a result, computational scientists are now seriously looking at potential uses of Java for scientic computing. But who is going to write scientic Java? Perhaps some numerical scientists will become procient in Java. But maybe we should also try to teach Java programmers the basics of numerical computing. Indeed, given the increasing scarcity of computer science graduates who are also trained in numerical analysis, perhaps this is a more realistic route to take. Ronald Maks book is very timely in this regard. It is targeted directly at beginning (or even not-so-beginning) Java programmers who would like to become familiar with numerical computing. It does not pretend to be an traditional course in numerical analysis, which quite frankly many present-day computer science undergraduates avoid like the plague. Rather, it teaches Java programmers what they need to know to be numerically literate, so as to be equipped to take on serious technical computing tasks when needed. The book starts out by describing in detail the IEEE-754 oating-point standard, both single and double formats. The author rst drives home the point that IEEE arithmetic is not the same as the real number systemfor example, there is potential for signicant loss of accuracy when two nearby oating-point values are subtracted. The book continues with topics such as the potential for diculties when a large number of oating-point values of dierent sizes are summed, nding roots of equations using basic iterative techniques such as Newtons iteration, nding interpolating and data-tting polynomials, and linear regression. Chapter 7, for example, discusses the trapezoidal rule and Simpsons rule for integration. It also presents techniques for numerical solutions to dierential equations, including Eulers method and the Runge-Kutta scheme. Again, the restraint that the author exercises here is remarkable. Im sure that most numerical analysts writing such a book could not resist the temptation to include here a furtive write-up of their favorite advanced 1

techniques, say for numerical quadrature. Space could also have been devoted to detailed discussion of techniques for numerical solutions of 2-D and 3-D partial dierential equations. Instead, Mak continues to stick to his formula of providing a very detailed and readable account of basic numerical methods. Beginning in Chapter 9, the author discusses matrix computations. Here, as in some previous chapters, the author provides a Java software package, which in this case is for matrix operations. With this facility, the author can focus on the concepts of matrix computation rather than on the detailed mechanics of carrying out such computations. Issues such as matrix condition numbers are discussed in Chapter 11. Even here, the focus is on concepts rather than on theorems, proofs or advanced implementation techniques. Part IV (beginning with Chapter 12), entitled The Joys of Computation, starts out by saying Numerical computation isnt all work and no play. In these chapters the author gives several examples of computations using a BigNumber and a BigDecimal package, which perform high-precision integer and oating-point computation, respectively. Some of us would choose to dier with Maks characterization of this material as play, since in recent years numerous important mathematical and scientic results have been obtained using such high-precision computations. Besides, much of our Internet commerce relies on security schemes based on high-precision arithmetic. In any event, this material is actually quite well written. The author covers topics such as large integer factorizations, computing mathematical constants and functions to high precision and fractals. In summary, this book would make an excellent undergraduate course in numerical computing, suitable for a wide range of students in computer science, physical science and engineering. It is also very well suited to professional Java programmers who would like to become more familiar with the world of scientic computing. It has a practical, down-to-earth approach that avoids exotic material, opting instead for a thorough and understandable coverage of basic material. It includes (and in fact relies on) software available from a website. Many students or other readers, after completing Maks book, will continue their careers with a greater appreciation of the issues and techniques of numerical computing, although perhaps they will not specialize in this arena. Others may nd the topic suciently engaging that they will pursue more serious coursework and career paths in scientic computing. Either way, the computational science community will be enriched as a result.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy