readme

亚定方程组求解：If serial correlation is found, you may have misspecified your model and should return to y

This is the lp_solve 2.3 release

bug reports, succes stories and requests for changes are welcome at:

Michel Berkelaar
[e-mail]

Most of the changes between 1.5 and 2.0 were contributed by:

Jeroen Dirks
[e-mail]

BUILDING LP_SOLVE

Just edit the Makefile to pick the compiler, lex and yacc of your choice, and
to set the desired compiler options. Then type 'make'. After a succesful make,
type 'make test' to run a couple of examples and to test the results against
the ones I got. Also run program 'demo' to test the procedural interface.


BUILDING LP_SOLVE for MS-DOS systems

Unfortunately the makefile.dos file its native for Borland compiler, since I've not
much practice with makefile, and it won't probably work correctly with other compilers.  
But you can still use it just as a trace for understanding how file are subdived and
which are the targets.
For the lucky ducks using BC++ 5.0 (or higher), the project file LP23.IDE can be used.
FLEX.EXE can be used for LEX and BISON.exe for YACC 
(downloadable at http://www.monmouth.com/~wstreett/lex-yacc/lex-yacc.html).  
Anyway the resulting file LPTAB.C is also provided.
Remember to use the "ANSI ONLY option" when compiling.
All the EXE and LIB 32 bit file generated are included.
Test examples can be run using:  
LP_SOLVE.EXE < example.lp > example.tmp 
and the comparing the results contained in example.tmp with that in example.out