igorcantele/optimization

optimization's Issues

Don't know how to integrate BMG with DEAP

optimization/optimizer/test-deap.py

Lines 11 to 16 in 807626f

	toolbox.register("attr_bool", random.randint, 0, 1)
	# Structure initializers
	toolbox.register("individual", tools.initRepeat, creator.Individual,
	toolbox.attr_bool, 100)
	toolbox.register("population", tools.initRepeat, list, toolbox.individual)

@Helveg

The example generates from random.randint a list of 100 bools and is optimized to generate empty list.
Individuals are the numbers to be optimized (the error of the BMG activity), while the population is the group of the individuals.

Two main problems:

• Here attr_bool is the random function we are using to fill the 100-lengthy list that will be our population. How to transform it in the double input of the function? I guess we can generate a np.random.random(2) and transform the free params in 0-1 values.
• I couldn't find any way to return the input of the function, I can only return the best Individual. I couldn't make an individual a dict, I guess I can make the first element of the Individual the pair of free params and the others the errors to sum up.