Linux binary available
- Clone the repository
git clone https://github.com/tveness/spinchain.git - Ensure the rust toolchain is installed, see https://rustup.rs/
- Run
cargo build --releaseto compile a fast version of the code, it may take a couple of minutes - The binary is now located at target/release/sc
sc --helpwill product the help dialogue listed belowscwill run a simple chain, and write the default configuration toconfig.toml
Classical spin-chain simulation for project investigating periodic driving of a spin chain coupled to a bath.
Numerically simulates the time-evolution of the classical spin-chain given by
$$
H = - \sum_j {\bf S_j} J_j {\bf S_{j+1}} + \sum_j {\bf B_j} \cdot {\bf S_j}
$$
where
sc 0.1.36-rc6
Thomas Veness <[email protected]>
Run classical spin chain simulation
USAGE:
sc [OPTIONS]
OPTIONS:
-a, --avg
Calculate average of runs
-A, --adiabatic <TAU>
Find adiabatic ensemble H-\\omega S^z
-b, --beta <BETA>
Override config beta
--config-desc
Print description of config file
-d, --dynamic-histogram [<POINTS>]
Generate time-evolution histogram (hist_dyn.dat) (default POINTS=1000)
-D, --adiab_full <TAU>
Find adiabatic ensemble H-\\omega S^z inhomogeneous
-e, --ext
Produce profile of energy density as a function of t
-E, --magnus-fit <E>
Fit temperature of Monte-Carlo ensemble with energy density E in lab frame, leading
Magnus
-f <E>
Fit temperature of Monte Carlo ensemble with energy density E
-F <E>
Calculate effective ensemble temperature via conserved quantity arguments
-h, --histogram [<POINTS>]
Generate histograms via Monte-Carlo (hist_mc.dat) and via time-evolution (hist_dyn.dat)
(default POINTS=8000)
-H, --high-freq <E>
Calculate effective ensemble with initial temp and first-order Magnus
--help
Print help information
-l, --langevin [<GAMMA>]
Directly simulate Langevin dynamics on the system proper (default GAMMA=1.0)
-m, --monte-carlo
Calculate an average quantitiy in Monte-Carlo
-m, --monte-carlo-magnus
Calculate an average quantitity in Monte-Carlo, Magnus expansion to second order in
omega^{-1}
--magnus-hist [<POINTS>]
Generate histogram via Monte-Carlo (hist_mc_magnus.dat) for first-order Magnus
expansion, and print averages (default POINTS=8000)
-n, --n-steps <TAU1,TAU2,...>
Generate single-shot time-evolution at different drive periods tau
-p, --monte-carlo-profile
Calculate an average quantitiy in Monte-Carlo, spatially resolved
-P, --dynamic-profile
Calculate an average quantitity in dynamical runs, spatially resolved
-r, --rot-frame <E>
Fit temperature of Monte-Carlo ensemble with energy density E in rotating frame
-s, --trajectory [<POINTS>]
Generate time-evolution histogram (hist_sj.dat) (default POINTS=1000)
--steps <STEPS>
-t, --tau <TAU>
Override tau from config
-V, --version
Print version information
In config.toml (if not present, default will be generated when running), there
are the following parameters (obtains by running sc --config-desc)
# Default config and description of options
hsize = 512 # size of the system
ssize = 8 # size of subsystem (driven part)
t = 256 # final time of simulation
dt = 0.02 # time-step for simulation
runs = 2 # number runs to perform
threads = 2 # number of parallel threads
trel = 0 # minus initial time
tau = 10 # period of drive
lambda = 1 # value of J_z coupling
hfield = [0.0, 0.0, 0.0] # constant h field on entire system
hs = [0.0, 0.0, 0.0] # constant h field on subsystem
jvar = 0.001 # variance in J couplings (x, y, z independent)
hvar = 0 # variance in field
method = 2 # method for numerical integration (2=2nd order Suzuki-Trotter)
ednsty = -0.66 # energy-density of initial state
file = "log" # pattern for log files i.e. log0.dat, log1.dat
strob = false # stroboscopic evaluation
offset = 0 # first file i.e. log0.dat
beta = 2.88 # beta (determined from ednsty)
e = 0.9 # eccentricity (scale of y-drive if drivetype="xyelliptic")
drivetype = "xyplane" # type of driving, can be "xyplane", "uniaxial", "xyelliptic"
By default when run, the program will produce a set of time-evolved data from an initial state.
It will log energy density of entire system, and then energy density for the subsystem, as well as subsystem magnetisations.
Comparing histograms for Monte Carlo and dynamical evolution to check that the system thermalises correctly. Plots here for beta = 2.89, and B-field = [1.0,0.0,0.0]
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent