Non optimalité corrigé dans individu evaluate

algogenetique.py

@@ -13,6 +13,7 @@ import time
 # Debut du decompte du temps
 start_time = time.time()
 
+#Les tests de temps ont été effectué avec 200 personnes, 50 survivant, 0.01 prob muté, 10 gen
 
 def main(N,tmax,pmutation, proportion):
     
@@ -21,21 +22,20 @@ def main(N,tmax,pmutation, proportion):
     brin = ''.join(lineList[1:])
     People=Population(N)
     S1=[]
-    for individu in People.indiv:
-        individu.evaluate()
+    for individu in People.indiv: #la boucle prend 8 secondes à faire en tout
+        individu.evaluate() #chaque appel coute 0.05 secondes 
         S1.append(int(individu.score))
     maximum=int(max(S1))
     for i in range(tmax):
         mini=People.indiv[0].score
         best=People.indiv[0]
-        People.reproduction(p = proportion, proba_mutation= pmutation)
+        People.reproduction(p = proportion, proba_mutation= pmutation) #La reproduction prend 7 secondes 
         for individu in People.indiv:
             if individu.score<mini:
                 best=individu
                 mini=individu.score
         L.append(mini)
         print(i,":",mini)
-
     plt.subplot(221)
     plt.plot([i for i in range(tmax)], L)
     
@@ -56,7 +56,7 @@ def main(N,tmax,pmutation, proportion):
 
 lineList = [line.rstrip('\n') for line in open("plasmid_8k.fasta")]
 brin = ''.join(lineList[1:])
-best,People = main(200,10,0.01,100)
+best,People = main(200,10,0.01,50)
 test = Traj3D()
 test.compute(brin, best.table)
 test.draw("first_plot")

individu.py

@@ -3,6 +3,7 @@ from Traj3D import Traj3D
 import numpy as np
 from math import sqrt, inf
 from random import random, randrange
+import time
 
 P1 = 0.015
 
@@ -15,33 +16,33 @@ class Individu():
         lineList = [line.rstrip('\n') for line in open("plasmid_8k.fasta")]
         self.brin = ''.join(lineList[1:])
         #self.brin = "AAAGGATCTTCTTGAGATCCTTTTTTTCTGCGCGTAATCTGCTGCCAGTAAACGAAAAAACCGCCTGGGGAGGCGGTTTAGTCGAA"
-        self.score = self.evaluate()
-    def evaluate(self):
+        self.score = None
+    def evaluate(self): #0.05 sec par individu
         traj = Traj3D()
 
         numb_ajout = 6
 
         fisrt_seq = self.brin[0:numb_ajout]
         last_seq = self.brin[-numb_ajout:]
+        traj.compute(last_seq + self.brin + fisrt_seq, self.table) #0.015
 
-        traj.compute(last_seq + self.brin + fisrt_seq, self.table)
-        traj_array = np.array(traj.getTraj())
+        test_start = time.time()
+        traj_array = traj.getTraj() #Instantannée
+        test_end = time.time()
+        print(test_end - test_start)
 
         list_distance = []
 
-        begining = traj_array[0:2*numb_ajout, 0:3]
-        end = traj_array[-2*numb_ajout:, 0:3]
-
-        for i in range(numb_ajout):
+        begining = traj_array[0:2*numb_ajout]
+        end = traj_array[-2*numb_ajout:]
+        for i in range(numb_ajout): #Instantannée
 
                 nuc_coordonate_beg = begining[i]
                 nuc_coordonate_end = end[i]
                 distance_nuc = np.linalg.norm(nuc_coordonate_beg - nuc_coordonate_end, ord=2)
                 list_distance += [distance_nuc]
 
-
         self.score = max(list_distance)
-
         #return max(list_distance)
 
 
@@ -79,24 +80,6 @@ class Individu():
                     else :
                         #sur l'axe z il y a un moins
                         table_rotations[doublet2][coord] = - table_rotations[doublet][coord]
-        table_rotations = self.table.rot_table
-        for i in range(0,number_of_mutations):
-            tir = random()
-            if tir < proba :
-                doubletNumber = randrange(0,8)
-                counter = 0
-                for doublet in table_rotations:
-                    if counter==doubletNumber:
-                        break
-                    counter+=1
-                for coord in range(3):
-                    table_rotations[doublet][coord] =np.random.uniform(low = self.table.orta()[doublet][coord] - self.table.orta()[doublet][coord + 3], high = self.table.orta()[doublet][coord] + self.table.orta()[doublet][coord + 3])
-                    doublet2 = self.table.corr()[doublet]
-                    if coord == 0 or coord == 1 :
-                        table_rotations[doublet2][coord] = table_rotations[doublet][coord]
-                    else :
-                        #sur l'axe z il y a un moins
-                        table_rotations[doublet2][coord] = - table_rotations[doublet][coord]
 
 # individu1 = Individu(RotTable())
 # print(individu1.table.rot_table)

population.py

@@ -3,6 +3,7 @@ from individu import Individu
 from RotTable import RotTable
 from croisement import croisement_un_point, croisement_deux_points
 import copy
+import time
 
 class Population:
     def __init__(self,n):
@@ -154,19 +155,19 @@ class Population:
         if p == None :
             p = (self.n)//2
         vieille_taille = self.n
-        selection(p)
-        newself = [element for element in self.indiv]       
-        while len(newself)<vieille_taille:
+        selection(p) #Instantannée
+        newself = [element for element in self.indiv]   
+        while len(newself)<vieille_taille: #Cette boucle prend 6-7 sec
             m=randrange(0,self.n)
             t=randrange(0,self.n)
             x=copy.deepcopy(newself[m])
             y=copy.deepcopy(newself[t])
             couple_enfant = enfant(x,y)
-            for child in couple_enfant :
+            for child in couple_enfant : #Cette boucle prend 0.1 sec
                 lineList = [line.rstrip('\n') for line in open("plasmid_8k.fasta")]
                 brin = ''.join(lineList[1:])
-                child.mutation(proba_mutation)
-                child.evaluate()
+                child.mutation(proba_mutation) #Instannée
+                child.evaluate() # 0.05 par child <-- C'est ce qui ralenti le tout
             newself.append(couple_enfant[0])
             newself.append(couple_enfant[1])
         self = self.modifier_population(newself)