isingm 1.0.3

A pythonic Ising model simulation

Ising

A pythonic implementation of the Ising model.

Check the example for more.

Lattice

This is the main place of the Ising model, we have to initialize it with a tuple of the lattice shape that can be of any dimension or size. The lattice can be initialized with a random state (you can choose the state ratio up) or all in a state 1 or -1.

arguments:

• shape:

type: tuple(int)

The shape of the lattice is tested for 1 to 4 dimension. Every values of shape must be > 0

Optionnal:

• all:

type: int == 1 or -1

Set the state at all 1 or -1 it overpass the r parrameter.

• r:

type: float in [0, ..., 1]

Set a random state with a ratio r of 1 in the lattice (default r = 0.5).

• adj:

type: numpy.array

A vector of vector: is the representation of the spin interaction.

0	J	0
J	#	J
0	J	0

Will be written as [[1,0],[-1,0],[0,1],[0,-1]]

As default it's the left right up down direct neighbor matrix will be genereted (whatever this mean in 4 or more dimensions).

• J:

type: numpy.array or float

Is the interraction between spins, if J is an array he as to be the same length than adj.

(you can choose to make anisotropic iteractions !!)

• B:

type: numpy.array or float

Is the magnetic field imposed on the lattice, if B is an array he as to have the same shape of the lattice

• beta:

type: float

Beta is 1/(Kb * T) with T the absolute temp, and Kb is the Boltzmann constant.

Methods

randomize(self, ratio=0.5):

Randomize th lattice state with a given ratio of up state.

arguments:

Optionnal:

• ratio:

type: float in [0, ..., 1]

The ratio of up state.

all(self, state):

Set all the lattice to the same state.

arguments:

• state:

type: int == -1 or 1

Value of the spin site.

mag(self):

Compute the magnetization of the lattice.

returns:

• Magnetization

type: float

Magnetization of the lattice.

hamiltonian(self):

Compute the Halmitonian of each spin.

returns:

• local_Hamiltonian:

type: numpy array

The Hamiltonian calculated for each spin.

H(self):

Compute the Hamiltonian of the lattice.

returns:

• Hamiltonian

type: float

Hamiltonian of the lattice.

mH(self):

Compute the mean value of the Hamiltonian.

returns:

type: float

Mean value of the Hamiltonian.

get_state(self):

Method to get the state of the lattice.

returns:

• state:

type: numpy array

Return the copy of the state.

set_state(self, state):

Method to set the state of the lattice.

arguments:

• state:

type: numpy array

The new state.

Metropolis.algorithm

Is the class who solve the Ising model with the Metropolis algorithm

Implementation of the Metropolis algorithm

arguments:

• for the arguments look the lattice class.

Methods

step(self, n=0):

step apply the metropolis algorithm on n spin once

arguments:

Optional:

• n:

type: int

n must be strictly positive is the size of the sample