Framework to compute Schrödinger functional correlation functions based on LatticeGPU (https://igit.ific.uv.es/alramos/latticegpu.jl)
Find a file
Fernando P. Panadero 09f086c07d Readme file
2024-02-29 12:29:46 +01:00
cnfg Working version 2024-02-22 13:32:25 +01:00
input Log file and gauge measurements 2024-02-22 12:24:42 +01:00
output Minor changes. 2024-02-27 09:20:05 +01:00
src Minor changes. 2024-02-27 09:20:05 +01:00
.gitignore Working version 2024-02-22 13:32:25 +01:00
Manifest.toml Added enviroment. Stable version. 2024-02-23 16:47:10 +01:00
Project.toml Added enviroment. Stable version. 2024-02-23 16:47:10 +01:00
README.md Readme file 2024-02-29 12:29:46 +01:00
sfcf.jl Added enviroment. Stable version. 2024-02-23 16:47:10 +01:00

Readme sfcf.jl

This is a framework to compute Schrödinger functional correlation functions based on LatticeGPU (https://igit.ific.uv.es/alramos/latticegpu.jl).

Requirements

The required packages are listed below

  • LatticeGPU
  • CUDA
  • TOML
  • TimerOutputs
  • ArgParse
  • InteractiveUtils
  • BDIO
  • MD5

The versions are resolved with the enviroment in the main sfcf.jl directory. The recomended version of julia is julia 1.8.5.

Input file

A template for the input file can be found at input/sfcf.in.

  • [Space]
  • size : Lattice size. The last component is the time extent.
  • blocks : Sub-lattices parallelization on the GPU. Each sub-lattice length must divide each lattice length.
  • phi0 : Phases defining the field at the boundary t = 0.
  • phiT : Phases defining the field at the boundary t = T.
  • cG : Boundary improvement coefficient for the gauge field.
  • [Fermion]
  • beta : Bare coupling.
  • kappa : Value of the hopping parameter.
  • theta : Phase for the fermion field in the spatial directions.
  • csw : Sheikholeslami-Wohlert coefficient
  • ct : Boundary improvement coefficient for the fermion field.
  • [Solver]
  • tolerance : Stopping criteria for the solver. The residue of the solver is meassured as the normalized residue for a source given by gamma_5 * D_w * gamma_5 * B , where B is the boundary source for each spin/color.
  • maxiter : Maximum number of iterations for the solver.

Configuration input

Current allowed input formats for the configuration are:

  • Native LatticeGPU (default) : Native io format from LatticeGPU
  • cern format : Read the format output given the function export_cnfg_cern() in sfqcd (https://gitlab.ift.uam-csic.es/alberto/sfqcd). This is specified with the option --cern

Execution command

The available options are:

  • -c (needed) : Specifies the gauge configuration.
  • -i (needed) : Specifies the input file.
  • --cern : Read the gauge config with cern format

An example of an execution command is

$ julia1.8 sfcf.jl -i ./input/sfcf.in -c ./cnfg/randfield

note that the main program sfcf.jl assumes that ./ is the main sfcf.jl directory to activate the enviroment.

Output

The main program outputs two files. the name is given by the input file [Run] [name]:

  • name.log : Log file with differ/ent information about the run

  • name.bdio : BDIO file with the correlation functions. The BDIO.jl package can be found at https://gitlab.ift.uam-csic.es/alberto/bdio.jl . The structure is the following

    • Record 1 : Information about the parameters

    • Record 14 : Format ihdr check

    • Record 8 : Gauge config name and MD5 saved as String. Results for the correlation functions, saved in the following order: f_P, f_A, f_1, g_P, g_A, k_V, l_V, k_1, k_T, l_T. All of them are saved as Complex{Float64}

Note that if the BDIO file exists, the results will be appended in Record 8.