import numpy as np
"""
f(x) = x*x + y*y - 5
"""
# Initializing n = 6
n = 6
chromosome = np.random.randint(0,5,(n,2))
# for num in chromosome:
#   for vl in num:
#     vl*=1.0
print("chromosomes :",chromosome)
epoch = 0

prev = 0;

while epoch <  200 :
    objective = abs(5 - chromosome[:,0]**2 - chromosome[:,1]**2 )
    avg = objective.sum()/n

    if abs(avg-prev)<1e-7 :
      break
    prev=avg

    print("Fitness object :", objective)
    print("Avg: ",avg)

    # Selection of fittest chromosome
    fitness =  1/(1 + objective)
    print("Fitness :",fitness)

    # Calculating the total of fitness function
    total = fitness.sum()
    print("Total :",total)



    # Calculating Probablility for each chromosome
    prob = fitness/total
    print("Probability :",prob)

    # Selection using Roulette Wheel And Calculating Cumulative Probability
    cum_sum = np.cumsum(prob)
    print("Cumulative Sum :", cum_sum)

    # Generating Random Numbers in the range 0-1
    Ran_nums = np.random.random((chromosome.shape[0]))
    print("Random Numbers :",Ran_nums)

    # Making a new matrix of chromosome for calculation purpose
    chromosome_2 = np.zeros((chromosome.shape[0],2))

    for i in range(Ran_nums.shape[0]):
        for j in range(chromosome.shape[0]):
            if Ran_nums[i]  < cum_sum[j]:
                chromosome_2[i,:] = chromosome[j,:]
                break

    chromosome = chromosome_2
    print("Chromosomes after updation :",chromosome)

    # crossover
    R = [np.random.random() for i in range(n)]
    print("Random Values :",R)

    # Crossover Rate
    pc = 0.25
    flag = Ran_nums < pc
    print("Flagged Values :",flag)

    # Determining the cross chromosomes
    cross_chromosome = chromosome[[(i == True) for i in flag]]
    print("Cross chromosome :",cross_chromosome)
    len_cross_chrom = len(cross_chromosome)

    # Calculating cross values
    # cross_values = np.random.randint(1,3,len_cross_chrom)
    # print("Cross Values :",cross_values)

    cpy_chromosome = np.zeros(cross_chromosome.shape)

    # Performing Cross-Over

    # Copying the chromosome values for calculations
    for i in range(cross_chromosome.shape[0]):
        cpy_chromosome[i , :] = cross_chromosome[i , :]

    if len_cross_chrom == 1:
        cross_chromosome = cross_chromosome
    else :
        for i in range(len_cross_chrom):
            c_val = 1
            if i == len_cross_chrom - 1 :
                cross_chromosome[i , c_val:] = cpy_chromosome[0 , c_val:]
            else :
                cross_chromosome[i , c_val:] = cpy_chromosome[i+1 , c_val:]

    print("Crossovered Chromosome :",cross_chromosome)

    index_chromosome = 0
    index_newchromosome = 0
    for i in flag :
        if i == True :
            chromosome[index_chromosome, :] = cross_chromosome[index_newchromosome, :]
            index_newchromosome = index_newchromosome + 1
        index_chromosome = index_chromosome + 1

    print("New Chromosomes:", chromosome)

    # Calculating the total no. of generations
    a ,b = chromosome.shape[0] ,chromosome.shape[1]
    total_gen = a*b
    print("Total Generations :",total_gen)

    #mutation rate = pm
    pm = 0.1
    no_of_mutations = int(np.round(pm * total_gen))
    print("No. of Mutations :" ,no_of_mutations)

    # Calculating the Generation number
    gen_num = np.random.randint(0,total_gen - 1, no_of_mutations)
    print(" Generated Random Numbers : " , gen_num)

    # Generating a random number which can replace the selected chromosome to be mutated
    Replacing_num = np.random.randint(0,5, no_of_mutations)
    # for num in Replacing_num:
    #   num*=1.0

    print(" Numbers to be replaced : " , Replacing_num)

    for i in range(no_of_mutations):
        a = gen_num[i]
        row = a//2
        col = a%2
        chromosome[row , col] = Replacing_num[i]

    print(" Chromosomes After Mutation : " , chromosome)

    epoch = epoch + 1