1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
|
INTRODUCTION
------------
ZIMPL is a little language to translate the mathematical model of a
problem into a linear or (mixed-)integer mathematical program
expressed in .lp or .mps file format which can be read by a LP or MIP
solver.
If you use Zimpl for research and publish results, the best way
to refer to Zimpl is to cite my Ph.d. thesis:
@PHDTHESIS{Koch2004,
author = "Thorsten Koch",
title = "Rapid Mathematical Programming",
year = "2004",
school = "Technische {Universit\"at} Berlin",
url = "http://www.zib.de/Publications/abstracts/ZR-04-58/",
note = "ZIB-Report 04-58",
}
COMPILING
---------
To compile Zimpl you need two additional librarys:
- GNU Multiple Precision Arithmetic Library
Version 4.2.2 or newer at http://www.swox.com/gmp
- zlib Version 1.2.5 or newer at http://www.gzip.org/zlib/
Zimpl has not been tested with older versions of above librarys.
GMP Version 3.x will not work for sure.
If zlib.h and gmp.h and libgmp.a and libz.a are not at the usual
places you have to adapt the Makefile, see below.
On a Linux system you might succeed with just saying: make
Otherwise:
Whatever with GNU C : gmake
Linux/x86 with Intel C : gmake COMP=intel
Tru64 with Compaq C : gmake COMP=compaq
Solaris with SUN Forte C : gmake COMP=sun
AIX with VisualAge C : gmake COMP=ibm
Windows with Cygwin/GCC : gmake COMP=gnu
HPUX with HP C : gmake COMP=hp
IRIX with SGI C : gmake COMP=sgi
If your configuration is not in the list or does not work,
you have to change the Makefile. Go to make/local
an add a file make.<hostname_of_your_computer>
There you can set the usual suspects like CC, CFLAGS, LDFLAG,
and so on.
The most likely reason that you can't compile, is that
gmp.h or zlib.h are not in your include path (-I) or
that libgmp.a oder libz.a are not in your library path (-L).
Example:
CPPFLAGS += -I/tmp/gmp/include
LDFLAGS += -L/tmp/gmp/lib
Have a look at Makefile and make/make.* to
get an idea.
On some MAC systems, GMP is installed under /sw/include and /sw/lib.
If these are not contained in the library and include paths, you have
to add them explicitly, as shown above.
WINDOWS
-------
You can just the through the source files into Visual Studio 2010.
Please add the contents of the src/WIN directory to the files to be
compiled. Make sure to set the following defines:
WIN32 or WIN64 and INLINE=
As for the libraries we recommend to replace gmp by
MPIR - Multiple Precision Integers and Rationals
version 2.4.0 or newer from http://mpir.org
Furthermore, you need
PCRE - Perl Compatible Regular Expressions
version 8.12 or newer from http://www.pcre.org
which should be compiled using the following two
defines -DWITH_PCRE -DPCRE_STATIC
CHECKING
--------
Do a "make check" and look if it works. If not, compile again
with an added "OPT=dbg" to the gmake invocation.
If a failure in the 'subto' test is reported, this is to a
difference in the behaviour of printf() and not a problem.
INSTALLATION
------------
Copy bin/zimpl.*.opt to wherever you like.
ALL GNULESS
-----------
if you don't have gcc and gmake etc. You can try the following.
Go to src and say "cc -O *.c -o zimpl".
LICENSE
-------
ZIMPL is under the GNU Lesser General Public License Version 3.
See the file LICENSE for details.
LATEST VERSION
--------------
Can be found at http://zimpl.zib.de
BUGS
----
Probably several. If you find one, send a description
with runnable example .zpl file to koch@zib.de
(Better even, send a working fix ;-)
REMARK
------
If anything is changed in the code I would recommend recompiling
without NDEBUG and NO_MSHELL defined. This will slow down the code
considerably but will do a lot of consistency checking.
In the source code are a lot of comments starting with
/*lint ... and also a target in the Makefile.
These are for flexelint (see http://www.gimpel.com) a very good
version of lint. If you are interested in my settings used for this
project, send me an email.
THANKS
------
To Tuomo Takkula, Armin Fuegenschuh, Marc Pfetsch, Daniel Junglas,
Joerg Rambau, Tobias Achterberg, Andreas Bley, Sebastian Orlowski,
Bejamin Hiller, Ralf Borndoerfer, Martin Assmann, Nicolas Weber,
Marc Moerig, Jens Hillmann, Dimitri Sverdlov, Joachim Reichel,
Maciej Warszawski and all I forgot for their comments and bug reports.
COMPLAINTS
----------
>/dev/null
Have fun!
|