This is a duplicate of README.txt
University of Southern Denmark
_____________________________________________________
Particle mass estimation using jackknife resampling
by krlor17
vibra17, heroe17, jakal17, ishac16
_____________________________________________________
Acknowledgements: This project was made possible through the guidance of
Benjamin Jäger
John Bulava
As we'd otherwise not know what jackknife resampling was in the first place.
--- Purpose of script ---
The matlab script DataJK.m naively estimates the masses of the 11 particles in the data/ directory.
The data are values of the time dependent correlator from a QCD sim.
The script employs Jackknife resampling to estimate statistical error
- without the bother of considering error propagation.
--- Running the script --- A MATLAB installation is required. Matlab R2017b or newer is recommended.
If possible, running the script in an environment not using a x.org server makes the process quicker (e.g. SSH connection to a ubuntu server system), as MATLAB will then select SOFTWARE OPENGL rendering. (see bottom) In a x server environment (or equivalent) MATLAB will run JVM by default, putting quite a load on low-performance systems.
It is not possible to run the script using the -nojvm option for MATLAB.
-- Linux --
To run the script for all particles on linux, simply execute run.sh $ ./run.sh
Make sure that the script has the needed permissions i.e. $ chmod 770 run.sh
-- Windows --
The easiest approach is to call DataJK(datafile, name) from the MATLAB IDE. The $name gives the folder and file name under the generated results. I.e.
DataJK pion_ud_correl_all_data_ascii.dat pion
Makes the script save the results for the pion under results/pion/ i.e. the final results in .tex format will be saved as
" results/pion/pion_final_table.tex "
-- Mac --
Whatever you feel like.
-- Tips for running script via. ssh --
Cloning the git repository is an easy way to download the script to the host server.
$ git clone https://github.com/krlor17/proj13_FF501_SDU.git