ifm is a simple and efficient implementation of the Wolff cluster algorithm
(Wolff, 1989)
for Monte Carlo simulation of Ising ferromagnets. It is optimized for very
large lattices and can handle billions of spins.
Simulation parameters are defined as macro constants to allow for effective
compiler optimization. Use the script setup.sh to prepare a build for a given
set of parameters. Binaries with different sets of parameters should be built
in separate directories.
For example, to build a simulation of a 2-d model on a 64x64 square lattice with periodic boundary conditions:
mkdir -p build/d2-L64
cd build/d2-L64
BC=BC_PERIODIC LT_D=2 LT_LMAX=64 ../../setup.sh
makeThe ifm binary takes a single optional argument, which is used to seed the
random number generator. Temperatures are read from standard input. Typical
usage looks like
./ifm 123 < temps.txt > d2-L64-123.datThe following environment variables are used by setup.sh:
Basic parameters
| Parameter | Type | Description |
|---|---|---|
BC |
{BC_PERIODIC,BC_FREE} |
boundary conditions |
LT_D |
integer | dimension of lattice |
LT_LMAX |
integer | linear size of lattice |
LOG_TIME_THERM |
integer | log time to thermalize |
LOG_TIME_MEAS |
integer | log time to measure |
Advanced parameters (usually don't need adjustment)
| Parameter | Type | Description |
|---|---|---|
RNG |
[see rng.h.in for values] |
type of random number generator to use |
UPDATES_PER_STEP |
integer | number of cluster updates per unit time |
MAX_STACK_SIZE |
integer | size of the stack used for cluster updates |