Lattice Models

Code for making Ising Model-like simulations using Metropolis-Hastings Algorithm and Wang-Landau algorithm.

Instalation

Code for simulations is written in Julia and plots and data analysis is made on Python. Dependencies can be checked on the |.toml files for Julia (which can also be used to build the project) and in REQUIREpython for Python.

For instalation, run install.sh file on a terminal. To test instalation, run test.sh

Documentation

All of the documentation regarding the functions can be checked in the sources files inside src directory.

Running

The scripts that run simulations are the src/run.jl and src/runDistributed.jl files. By default, src/runDistributed.jl tries to use all of the cores available in the machine to parallelize the simulation. For more information regarding their use, run them with the --help flag.

Outputs

All of the simulations are stored in a folder with the name of the date and time they were performed. Inside the each folder, there will be the following files:

name	descrition
adjMat.csv	Adjacency matrix of the network of the simulation
algoParam.csv	Parameters of the algorithm used (number of steps, name)
metaParam.csv	Parameters of the network, model the energy function
simulParamDict.csv	Table with all the values of the energy function parameters that were simulated and the order in which they were performed. Order is important for hystheresis analysis

The files beggining with i_j represent the j independent iteration of the simulation ran with the energy parameters of the i+1 csv of simulParamDict.csv . Their meaning is the following.

name	descrition
i_j_initial.csv	Linearized array of the sites values at the beggining of each simulation.
i_j_changes.csv	Linearized array representing the changes at each MC step. If k=0, no change was made at that step. If k!=0, the site k was changed in that moment
i_j_mag.csv,i_j_ener.csv	Energy and magnetization values in each time step

Warning

If the outputs folder hasn't been created, if is possible that the first run of src/runDistributed.jl will throw an error.