Skip to content

Simulation Outputs

Jump to bottom
Andrea Rucci edited this page Jan 31, 2018 · 4 revisions

The 1d_pimc executable will print on file the outputs in the working directory specified when the program was run. Three output files are produced, each one with an header in the first line (starting with #) in which all the information given in the init file are printed so that to avoid losing informations about the run that generated the data, but can be also useful if the data is used in next analysis. For example, the header corresponding to the init file in the Readme.md would be

# 4 100 0.1 1 100 harmonic 2 3.0 0. nointeraction 1 0.0 5000 10 0
data
data
data
...

Produced file are the following:

  • accpt is a one-column file in which the acceptance of a single update step is printed. Remember that, in the program, a single update is actually a series of Np*Nt Metropolis steps on each particle trajectory on each time-slice, so that the acceptance is just the number of time a random move is accepted divided by the number of trials.

  • meas is the collection of the measure on the system. In this version there is only the computation and print of the action of the system, so in the file you will find just a single column of numbers. If needed, you may add new functions in order to measure other observables (see the section Upgrade and Modify the Code). On the other hand, you can also use the configurations themselves in the other output files and compute the new observables in a second stage with other scripts and codes). Using only the action you can have an hint about how the simulation goes and if the thermalization of the system took place.

  • confs is the file containing the saved configuration of the system. Each line is a configuration printed following the structure explained in the Overview and Structure section. So, the number of columns would be Np*Nt, while the number of lines is the ratio between MCsteps and MCeach. The expected size of this file is printed on the screen while running the simulation (it should be about 1Mb for a collection of a thousand configurations with Nt around the hundreds and Np of the order of ten).

Clone this wiki locally