================================================================ Grace ver.2.2.1 2004/10/11 Minami-Tateya GRACE is a program package developed for automating calculations of cross sections in the perturbation theory of high energy physics. I. SYSTEM REQUIREMENT Installation of this system requires the following software on a Unix/Linux operating system: 1. C compiler 2. Fortran compiler 3. "sh", "make" and other standard Unix commands 4. X Window system (optional) 5. Motif Toolkit or its clone (optional) II. INSTALLATION 1. Getting source code Gziped tar file of Grace system is available from the following web page : http://minami-home.kek.jp/ 2. Expand source code Source code directory tree is expanded by the following command : tar xvzf grace.221.2004.1012.tar.gz with GNU-tar command or gunzip grace.221.2004.1012.tar.gz tar xvf grace.221.2004.1012.tar with Unix tar command. A new directory grace.2.2.1 and its subdirectories are to be created. 3. Create Makefile First move to the top of expanded grace directory tree cd grace.2.2.1 The command ./Config.sh creates "Makefile" with the default setting, which can be shown by make -n make -n install In order to change installation directory, compiler command or compiler options, it is necessary to edit this "Makefile". The meaning of the variables are described in the same file. 4. The programs of Grace system are compiled by make 5. Compiled code and runtime library is installed by make install into the directory specified in "Makefile". 6. Installation can be tested by make test-install This command checks the gauge invariance of the generated code for e+ e- --> W+ W- photon in the standard model and for e+ e- --> photon Chargino1 anti-Chargino1 in the MSSM model. The test running of the program shows the difference of the values of differential cross section calculated at one phase space point with different values of gauge parameters. In this calculation particle widths are set to zero in order to restore the gauge invariance. When the Grace system is installed successfully, the relative difference shown in this calculation should be less than 1E-30 in the quadruple precision. Subdirectories testinst/sm/eewwa/fort and testinst/mssm/asw1SW1/fort are created in this installation test. These subdirectories keep generated Fortran and compiled code. Here you will find the files named "*.log" and "gauge.out" which are the logs of commands and the output of gauge invariance check, respectively. III. SAMPLE CALCULATION You can perform the above test calculation in step by step by hand and continue to calculate total cross section and to generate simulated events. 1. Before calculation, you should add the installed binary directory to your command search path setenv PATH $PATH:installed_directory_path/bin rehash in "csh" or "tcsh", or PATH=$PATH:installed_directory_path/bin export PATH in "sh" or "bash". 2. Preparation of input file 2.1 Change directory to "testinst/mssm/asw1SW1" or "testinst/sm/eewwa/fort" cd testinst/mssm/asw1SW1 2.2 Make a working directory mkdir fort 2.3 Copy an input file prepared as "input/in.prc" to the working directory cp input/in.prc fort 2.4 Move to the working directory cd fort If you want to calculate another process, you should edit "in.prc". 3. Feynman graphs are generated by grc This command reads "in.prc" and creates a file "out.grf" which is a text file describing the structure of generated Feynman graphs. 4. When X Window and Motif toolkit have been found on your system at the installation of Grace system, drawing tools of Feynman graphs are to be installed. In this case, Feynman graphs can be shown on the screen by gracefig or grcdraw -h grcdraw 5. A set of Fortran code is generated by grcfort The generated code is compiled by make make integ make spring or make all 6. We recommend to confirm the gauge invariance of the differential cross section first by typing ./gauge This check is included in the standard procedure of installation test (see II-6). When a comment line CC call prmass found in "gauge.f" is uncommented, commands make ./gauge show the values of mass parameters and coupling constants used in this calculation. 7. The differential cross section is integrated over phase space by ./integ You should carefully examine the convergence behavior of of the integration. If the convergence is not so good, you must increase the number of sampling points or replace kinematics routines. 8. Using the data created in the previous step, one can generate simulated events by ./spring This command shows the distribution of generated events. The values of four vectors of the events are stored in array varaible "vec". If you want to keep them, you should modify "mainsp.f" so as to write the values of this array variable on a file in the suitable format to your purpose. IV. NOTICE In the case of MSSM, this package uses subroutines available from http://www.netlib.orig which are included in the subdirectory "netlib" of this distribution. These subroutines are written for the calculations of double precession. They may not work well in another precession.