readme.coool

COOOL：CWP面向对象最优化库（CWP Object Oriented Optimization Library） COOOL是C++类的一个集合

This is COOOL version 1.1. COOOL stands for the Center for Wave
Phenomena Object Oriented Optimization Library. The Center for Wave
Phenomena is an interdisciplinary research group at the Colorado
School of Mines engaged in research in seismic and acoustic modeling
and inversion, wave propagation, seismic data processing and
scientific computing.  Our WWW home page is at
http://timna.mines.edu/cwpcodes/coool/.

Comparing with the earlier version (COOOL 1.0), the new version
is tested on gcc-2.7.0. This version also works on the old
gcc-2.5.8. If you are using gcc 2.5.*, you can also ftp the
older version package COOOL-0.1.tar.gz.
There are some changing standards for the  new gcc compiler, so a
previous compiled c++ code may not compile under the new standard. 
Adapting the codes is not difficult but maybe tedious. Please refer 
to the new gcc texinfo files.  

The new COOOL can set up a symbolic link of SlaveForward file to the
appropriate file by detecting the machine type while installing. 
But you still need to read  Makefile.config to make appropriate
changed according to your system. A big improvement for this version
is that COOOL can take customized derivative information from the
user-defined file. Because the change of the current compiler, you
will see many warning messages. such as "conversion to the same type
will never use a type conversion operator ..." Ignore them!

This library is being developed not only to provide a set of tools for
solving optimization problems, but also to stimulate interest in
optimizers to look at the sorts of large-scale problems encountered in
geophysics.  The library has both derivative-based and derivative-free
optimization routines.  We wanted to be as flexible as possible in the
communication protocol connecting the library algorithms and the
user-coded objective and gradient function so we used EXPECT,  written
by Don Libes of NIST and freely available.  Besides the algorithms, we
provide C++ classes to manipulate vectors and (sparse or dense)
matrices that can be useful for scientific software development.

The following is a brief description of the optimization algorithms
that are available in the present version of the COOOL library:

Linear system solvers: 
	Conjugate Gradient for Least Squares (CGLS)
	Interactive Reweighted Least Squares (IRLS), which
		solves arbitrary linear systems in the l_p norm.

Non linear optimization tools:
	Derivative-free:
		Powell's method (Powell)
		Downhill Simplex (Simplex)
	Derivative-based:
		Conjugate Gradient (CG) with Polak-Ribiere and 
		Fletcher-Reeves	step computation procedures
		
Line searches:
	Derivative-free:
		Brent's algorithm (BrentLineSearch)
	Derivative-based:
		Cubic line search (CubicLineSearch)

Vector and Matrix manipulation classes:
	take a look at Vector.hh (Vector) and Matrix.hh (Matrix)

Man pages are available.  To get detailed information about the
algorithms listed above type man followed by whatever is between
parentheses (It is not case sensitive!).

For information on how to install the COOOL library and how to run the
demos look at the README.install file. For detailed information on the
internal structure of the library and about the optimization
algorithms check the postscript file available via WWW at the URL: 
	http://timna.mines.edu/cwpcodes/coool/
or via anonymous FTP to timna.mines.edu, "cd" to the directory
pub/cwpcodes/coool and transfer the binary file coool.ps.gz. You
should use gzip to uncompress this file.

The COOOL library is a free software. You can do anything you want
with it including make a fortune.  However, neither the authors,
the Center for Wave Phenomena, nor anyone else you can think of
makes any guarantees about anything in this package or any aspect
of its functionality.

Since you've got the source code you can also modify the
library to suit your own purposes. We would appreciate it 
if the headers that identify the authors are kept in the 
source code.

Email bug reports, questions, and suggestions to
        [e-mail].
We'll do our best to fix them as soon as possible.

Thanks for using COOOL.