Numerical Methods Framework
Book Back Cover
Inside Cover Chapter Map
EATSv5 File Load
All the code presented in this book is available in the accompanying CD ROM. The contents of the CD ROM is described in this file.
By using this CD-ROM you agree to be bound by the following agreement:
This software is supplied to the purchaser of the book for use for any purpose. This software is sold as is without any warranty of any kind, either expressed or implied, including but not limited to the implied warranty of merchantability and fitness for a particular purpose. Neither the author, nor Morgan Kaufmann, nor its dealers or distributors, assume any liability for an alleged or actual damages arising from the use, or the inability to use, this software.
The entire and exclusive remedy for any defect in materials or workmanship in construction of the CD-ROM shall be limited to replacement of the defective CD-ROM.
The Java code presented in this book was created with Visual Age for Java. The entire code is split into several packages: these packages correspond more or less to the chapters of this book. They differ slightly from the Smalltalk applications because the Java code was generated independently from the Smalltalk code. Also due to the fact that Java does not allow to extend a class in another file, all files contain only one class.
All classes defined in chapter 12.
- Cluster Analysis
- Covariance Analysis
- Mahalanobis Center
All classes defined in chapter 10.
- Paramaterized Function
- Polynomial Least-square Fit
- Linear Regression
- Non-linear Least-square Fit
- Maximum Likelihood Fit
- Chi-square test
All classes defined in chapters 1 and 2.
- Gamma function
- Beta function
- Error function
All interfaces defined in this book.
All classes defined in chapter 3.
- Newton interpolation
- Lagrange interpolation
- Cubic spline interpolation
- Neville interpolation
- Bulirsch-Stoer interpolation
All classes defined in chapters 4, 5, 6 and 7.
- Iterative process
- Function iterator
- Newton’s zero finder
- Bisection zero finder
- Trapeze integrator
- Simpson integrator
- Romberg integrator
- Infinite Series
- Continued function
- Incomplete gamma function
- Incomplete beta function
All classes defined in chapter 8.
- Symmetrical matrix
- LUP decomposition
- Jacobi eignvalues
All classes defined in chapter 11.
- Genetic algorithms
- Simplex minimization
The classes Histogram and Curve.
The classes defined in chapter 9, except class Histogram
and the classes defined in appendix D.
- Random number generator
- Statistical moments
- Probability distribution
Visual Age for Java
The file CDRom/Vaj/OONumJ.dat contains a Visual Age
repository. Users of Visual for Java can load the entire code by
loading this file using the import facility of Visual Age.
Other Java systems
Classes must be loaded from the hierarchy of directories rooted at
CDRom/Java. Each subdirectory corresponds to one of the packages
Note: Because some Java system requires that the name of the
class be identical to the name of the file, the original names
of the classes have been kept. Similarly, the name of the
directory must be equal to the name of the package. If you have
transferred the contents of the CD ROM over a system limiting the
size of the file name, you must first rename each file and
directory before loading the classes.
The directory Utility can only be used on a 32-bit Windows systems.
It contains two programs to study some numerical effects
experimentally. The directory must be copied as is onto a 32-bit
Windows systems (Windows NT 4.0 or higher, Windows 98 or higher).
The executable files must be run from this directory because it
contains all the resources needed to execute them.
The program contained in the file DistributionDemo.exe allows the
reader to study all probability density functions described in this
book. After starting the program, a window containing a note book
will open. Each page of the notebook corresponds to a distribution.
The parameters can be modified.
Then, clicking on the button labeled “Generate” will generate
random values distributed according to the distribution. A window
allows to modify the number of generated values and the parameters
of the histogram into which the values are accumulated. When the
generating is completed the resulting histogram is displayed. The
reader can then attempt a least square fit or a maximum likelihood
fit and view the result. The button labeled “Integral” allows
to view the values as a distribution function.
This small utility program allows to display data read from a text
file, either in tab or coma delimited format, in graphical form.
Once data are read, the user can investigate interpolation, linear
regression and polynomial least square fit. This program was used
to generate the figures of chapters 3 and 10.
After starting the program, a window containing the parameters of
the file containing the data will open. Here the reader can select
the name of the file to read, the format used (blank or coma
delimited) and which column contains which data. For error bars,
there are three cases:
- no error;
- the error is given explicitly;
- the error is computed from the standard deviation and the
number of values used to compute the standard deviation (cf
equation 9.6 in chapter 9).
If the data were read without error, the reader can investigate
the three interpolation algorithms described in chapter 3. In
addition linear regression and polynomial least square fit can be
tried, each point having an identical weight.
If the data were read with errors, the reader can investigate only
Lagrange and Bulirsch-Stoer interpolation. Linear regression and
polynomial least square fit are performed using the supplied error
to compute the weight of each point as described in chapter 10.