multiz Documentation



Evaluates 3pt functions and computes ratios of 3pt/2pt correlators. The output from the Mathematica script to generate the GPD parameterization is necessary. However, this version does no longer take recourse to separate executables to access the building block files. Hence, it is no longer compatible with QCDSF interface conventions.


For instructions, call "multiz" with the "-h" option.

Usage instructions


multiz [-v] {-b <building-blocks path> [-k <quarkmass>] -m <mathinp> \ [-t <tsink>] -i <tau_min> -a <tau_max> \ -s {d|i} -f {u|d|u-d|u+d|curr} [-j <Binsize>] \ [-2 <two-point function file name pattern>] \ [-3 <three-point function file name pattern>] \ [-c <configuration-file>]}

Compute the generalized form factors from a series of lattice simulations. The path to the building-block files is specified using the -b option. The quark mass (or, in case of Wilson fermions, the kappa-parameter) for the valence fermions is given using the -k option.

NOTE: The kappa-parameter is being used to normalize the quark states if the program has been compiled for Wilson fermions. Hence, it is obligatory to specify this parameter in that situation. The -m option specifies the Mathematica input file which contains the system of equations to relate operator matrix elements to the generalized form factors. This file has to be provided by an external program. To this end, we have a Mathematica program for this purpose, but the file is human-readable and can also be edited manually.

If the current version of Multiz supports an observable which you are interested in, you can provide it with an input file (see the -c option below) which describes the working point.

NOTE: This functionality is currently unsupported. Future versions may implement this feature if there is sufficient demand and on what exactly should be available.

The fitting range for the ratio R must be specified by the interval [tau_min,tau_max] which must satisfy
0 < tau_min <= tau_max < tsink.
The position of the sink is specified by the -t option. The system of equations is specified by the file <mathinp> which is usually generated automatically by Mathematica.

The -s flag indicates if you want to use compute spin-dependent (flag -s d) or spin-independent (flag -s i) GPDs. The user has to ensure himself that the appropriate Math-input file is selected for this flag. The -f flag indicates the requested quark flavor, which can be chosen to be either u, d, u-d, u+d, or curr for 2/3u-1/3d (which corresponds to the proton electromagnetic current). The extraction of n-point functions is done by accessing the building block files directly. No external programs are necessary anymore.

The optional switch -j specifies the Jackknife bin size. If omitted, it will default to 1.

The -v switch turns on verbose progress reporting.

Since there are varying filename conventions for the two- and three- functions and no consensus on a unified naming scheme, the pattern to be used can be set manually by a command line option. The -2 and -3 switches will change the naming convention and will take a C-style printf-format string for the appropriate options. However, for the exact format to be used the mio.c source file needs to be consulted.


Written 2000-2007 by LHPC collaboration W. Schroers <> MIT - Center for Theoretical Physics G.T. Fleming <> Yale University - Physics Department D.B. Renner <> University of Arizona - Department of Physics P. Haegler <> Universitaet Muenchen - Physik Department

 All Data Structures Files Functions Variables Enumerations Defines