frigidum 0.1.5

Simmulated Annealing

Simmulated Annealing package for Python

Installation

pip install frigidum

Example Usage

import frigidum

import random

def random_start():
    return 50 + random.random()

def random_small_step(x):
    return x + 0.1 * (random.random() - .5)

def random_big_step(x):
    return x + 10 * (random.random() - .5)

def obj(x):
    return x**2

local_opt = frigidum.sa(random_start=random_start, random_movements=[random_small_step, random_big_step], objective_function=obj, T_start=100, T_stop=0.001, repeats=10**4, copy_state=frigidum.annealing.naked)

Arguments:

• random_start : function which returns a random start / state
• objective_function : objective function to minimize
• random_movements : list of neighbor functions, for one use [func]
• t_start : Starting temperature
• T_stop : Stopping temperature
• alpha : lower temperature by this factor, after repeats proposals
• repeats : at each lowering by alpha, do repeats proposals
• copy = {'copy', 'deepcopy', 'naked'} - copy method

Movements

A movement is a when a proposed state is accepted, and the objective function has changed. For each batch of repeats, the proportion of movements are displayed.

In the early phase of annealing, movements should happen >90%.

In the last phase of annealing, movements should happen <5%.

Copy'ing of States

3 most important copy methods are included in the annealing module,

def copy(state):
	return state.copy()

def deepcopy(state):
	return state.deepcopy()

def naked(state):
	return state

In the example, naked with the argument copy_state=frigidum.annealing.naked is used,

• use copy_state=frigidum.annealing.copy for copy(),
• use copy_state=frigidum.annealing.deepcopy for deepcopy(),
• use copy_state=frigidum.annealing.naked if a = b would already create a copy.

General Advise with Simulated Annealing

• To get inspiration for random movements, try solve a similar problem yourself.
• Try/use multiple random movements, combinations usually work well.
• Try add movements, that might work well when cold.
• Try add movements, that might work well when warm.
• It is difficult to predict the effect of random movement, usually my ideas don't survive the outcome of experiments.
• When conditions apply, stay within the feasible zone when possible.

To-Do:

• Statistics of acceptance of various movements/neighbors
• Multitreadding (N simultanous anneals)
• Drilling (after repeats, re-repeat with low temp)
• Re-Annealing
• Auto-set start Tempreature (Based on >90% movemenets)
• Auto-stop (Based on near 0 movemebts)