import numpy as np 
import random 
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
#----------------------PSO參數設置--------------------------------- 
class PSO(): 
 def __init__(self,pN,dim,max_iter): #初始化類 設置粒子數量 位置信息維度 最大迭代次數 
  #self.w = 0.8 
  self.ws = 0.9
  self.we = 0.4
  self.c1 = 1.49445  
  self.c2 = 1.49445  
  self.r1= 0.6 
  self.r2= 0.3 
  self.pN = pN    #粒子數量 
  self.dim = dim    #搜索維度 
  self.max_iter = max_iter #迭代次數 
  self.X = np.zeros((self.pN,self.dim))  #所有粒子的位置（還要確定取值范圍） 
  self.Xmax = 5 
  self.Xmin = -5
  self.V = np.zeros((self.pN,self.dim))  #所有粒子的速度（還要確定取值范圍）
  self.Vmax = 1 
  self.Vmin = -1
  self.pbest = np.zeros((self.pN,self.dim)) #個體經歷的最佳位置 
  self.gbest = np.zeros((1,self.dim))   #全局最佳位置
  self.p_fit = np.zeros(self.pN)    #每個個體的歷史最佳適應值 
  self.fit = 0    #全局最佳適應值 
#---------------------目標函數Sphere函數----------------------------- 
 def function(self,x): 
  y = np.sin(10*np.pi*x)/x
  return y
 def Holder_table(self,x,y):  
  z = -np.abs(np.sin(x) * np.cos(y) * np.exp(np.abs(1 - np.sqrt(x**2 + y**2)/np.pi)))
  return z
 def fuck(self,x,y):
  z = x**2 + y**2 - 10*np.cos(2*np.pi*x) - 10*np.cos(2*np.pi*y) + 20
  return z
#---------------------初始化種群---------------------------------- 
 def init_Population(self):
  for i in range(self.pN):        #遍歷所有粒子
   for j in range(self.dim):       #每一個粒子的緯度
    self.X[i][j] = random.uniform(-5,5)    #給每一個粒子的位置賦一個初始隨機值（在一定范圍內）
    self.V[i][j] = random.uniform(-1,1)    #給每一個粒子的速度給一個初始隨機值（在一定范圍內）
   self.pbest[i] = self.X[i]       #把當前粒子位置作為這個粒子的最優(yōu)位置
   tmp = self.fuck(self.X[i][0],self.X[i][1])   #計算這個粒子的適應度值
   self.p_fit[i] = tmp         #當前粒子的適應度值作為個體最優(yōu)值
   if(tmp > self.fit):         #與當前全局最優(yōu)值做比較并選取更佳的全局最優(yōu)值
    self.fit = tmp 
    self.gbest = self.X[i] 
#---------------------更新粒子位置---------------------------------- 
 def iterator(self): 
  fitness = [] 
  for t in range(self.max_iter):
   w = self.ws - (self.ws - self.we) * (t / self.max_iter)
   for i in range(self.pN): 
    #更新速度
    self.V[i] = w*self.V[i] + self.c1*self.r1*(self.pbest[i] - self.X[i]) + self.c2*self.r2*(self.gbest - self.X[i])
    if self.V[i][0] > self.Vmax:
     self.V[i][0] = self.Vmax
    elif self.V[i][0] < self.Vmin:
     self.V[i][0] = self.Vmin
    if self.V[i][1] > self.Vmax:
     self.V[i][1] = self.Vmax
    elif self.V[i][1] < self.Vmin:
     self.V[i][1] = self.Vmin
    #更新位置
    self.X[i] = self.X[i] + self.V[i]
    if self.X[i][0] > self.Xmax:
     self.X[i][0] = self.Xmax
    elif self.X[i][0] < self.Xmin:
     self.X[i][0] = self.Xmin
    if self.X[i][1] > self.Xmax:
     self.X[i][1] = self.Xmax
    elif self.X[i][1] < self.Xmin:
     self.X[i][1] = self.Xmin
   for i in range(self.pN):   #更新gbest\pbest 
    temp = self.fuck(self.X[i][0],self.X[i][1]) 
    if(temp > self.p_fit[i]):  #更新個體最優(yōu) 
     self.pbest[i] = self.X[i]
     self.p_fit[i] = temp 
    if(temp > self.fit):   #更新全局最優(yōu) 
     self.gbest = self.X[i] 
     self.fit = temp 
   fitness.append(self.fit) 
   print('最優(yōu)值為：',self.fit)#輸出最優(yōu)值 
   z1 = self.fit
   print('最優(yōu)位置為：',self.X[i][0],self.X[i][1])
   x1 = self.X[i][0]
   y1 = self.X[i][1]
  return fitness, z1, x1,y1
#----------------------程序執(zhí)行----------------------- 
my_pso = PSO(pN=100,dim=2,max_iter=200) 
my_pso.init_Population() 
fitness,z1,x1,y1 = my_pso.iterator()
plt.figure(1) 
plt.title("Figure1") 
plt.xlabel("iterators", size=14) 
plt.ylabel("fitness", size=14) 
t = np.array([t for t in range(0,200)]) 
fitness = np.array(fitness) 
plt.plot(t,fitness, color='b',linewidth=3) 
plt.show() 
fig = plt.figure(figsize=(15,10))
ax = Axes3D(fig)
X = np.arange(-5,5,0.1)
Y = np.arange(-5,5,0.1)
X,Y = np.meshgrid(X,Y)
def f(x,y):
 return (x**2 + y**2 - 10*np.cos(2*np.pi*x) - 10*np.cos(2*np.pi*y) + 20)
ax.plot_surface(X,Y,f(X,Y),rstride=1,cstride=1,cmap= plt.get_cmap('rainbow'))
ax.scatter(x1, y1, z1,s=400,c='k',marker = '*')
plt.show()