Added genetic_algorithm.py file which implements genetic algorithm

algorithms/learn/genetic_algorithm.py

@@ -0,0 +1,133 @@
+import random
+from itertools import permutations
+
+# genetic algorithm search of the one max optimization problem
+from numpy.random import randint
+from numpy.random import rand
+
+# this is helper function for sum_distance function, it counts the taxi cab distance between 2 vectors [x,y]
+def distance(x,y):
+    temp1 = abs(x[0]-y[0])
+    temp2 = abs(x[1]-y[1])
+    vector_distance = temp1+temp2
+    return vector_distance
+
+# this is fitting function which tells how well specimen fits the environment
+# this function counts the sum of distances between vectors for a specimen
+# this was just for testing, it should be probably changed to A*
+def sum_distance(speciment):
+    sum = 0
+    for i in range(0,len(speciment)-2):
+        pom = distance(speciment[i],speciment[i+1])
+        sum = sum + pom
+    return sum
+
+
+# tournament selection
+# this function randomly puts speciments to groups, from each group one best speciment is taken for reproduction
+def selection(pop, scores, k=3):
+    # first random selection
+    selection_ix = randint(len(pop))
+    for ix in randint(0, len(pop), k- 1):
+        # check if better (e.g. perform a tournament)
+        if scores[ix] < scores[selection_ix]:
+            selection_ix = ix
+    return pop[selection_ix]
+
+
+# crossover two parents to create two children
+# this function creates speciments for new generation
+def crossover(p1, p2, r_cross):
+    p1=list(p1)
+    p2=list(p2)
+    # children are copies of parents by default
+    c1, c2 = p1.copy(), p2.copy()
+    # check for recombination
+    if rand() < r_cross:
+        # select crossover point that is not on the end of the string
+        pt = randint(1, len(p1) - 2)
+        # perform crossover
+        temp1 = p1[:pt]
+        for mine in p2:
+            if mine not in p1[:pt]:
+                temp1.append(mine)
+        temp2 = p2[:pt]
+        for mine in p1:
+            if mine not in p2[:pt]:
+                temp2.append(mine)
+        c1 = temp1
+        c2 = temp2
+    return [c1, c2]
+
+
+# mutation operator
+# this function checks whether genes mutated
+# if gene is mutated then it is swapped with randomly chosen gene from the same speciment
+def mutation(speciment, r_mut):
+    for i in range(len(speciment)-1):
+        # check for a mutation
+        if rand() < r_mut:
+            # flip the bit
+            temp = speciment[i]
+            pom = random.randint(0,len(speciment)-1)
+            speciment[i] = speciment[pom]
+            speciment[pom] = temp
+
+
+
+# genetic algorithm
+def genetic_algorithm(objective, n_iter, n_pop, r_cross, r_mut):
+    # this is hardcoded list of coordinates of all mines (for tests only) which represents one speciment in population
+    # it is then permutated to get set number of species and create population
+    speciment=[[1,1],[5,2],[1,3],[2,5],[2,1],[3,2],[3,4],[5,0]]
+    pop = [random.choice(list(permutations(speciment,len(speciment)))) for _ in range(n_pop)]
+    # permutation function returns tuples so I change them to lists
+    for i in range(len(pop)):
+        pop[i] = list(pop[i])
+
+    # keep track of best solution
+    best, best_eval = 0, objective(pop[0])
+
+    # enumerate generations
+    for gen in range(n_iter):
+        # evaluate all candidates in the population
+        scores = [objective(c) for c in pop]
+        # check for new best solution
+        for i in range(n_pop):
+            if scores[i] < best_eval:
+                best, best_eval = pop[i], scores[i]
+                print(">%d, new best f(%s) = %.3f" % (gen, pop[i], scores[i]))
+        # select parents
+        selected = [selection(pop, scores) for _ in range(n_pop)]
+        # create the next generation
+        children = list()
+        for i in range(0, n_pop, 2):
+            # get selected parents in pairs
+            p1, p2 = selected[i], selected[i + 1]
+            # crossover and mutation
+            for c in crossover(p1, p2, r_cross):
+                # mutation
+                mutation(c, r_mut)
+                # store for next generation
+                children.append(c)
+        # replace population
+        pop = children
+    return [best, best_eval]
+
+
+# define the total iterations
+n_iter = 100
+# bits
+n_bits = 20
+# define the population size
+n_pop = 100
+# crossover rate
+r_cross = 0.9
+# mutation rate
+r_mut = 0.05
+# perform the genetic algorithm search
+best, score = genetic_algorithm(sum_distance, n_iter, n_pop, r_cross, r_mut)
+print('Done!')
+print('f(%s) = %f' % (best, score))
+
+