python 梯度法求解函数极值的实例

            
#coding utf-8
a=0.001    #定义收敛步长
xd=1    #定义寻找步长
x=0     #定义一个种子x0
i=0     #循环迭代次数
y=0
dic={}
import math
def f(x):
 y=math.sin(x) #定义函数f(X)=sinx
 return y
def fd(x):
 y=math.cos(x) #函数f(x)导数fd(X)=cosx
 return y
while y>=0 and y<3.14*4:
 y=y+xd
 x=y
 while abs(fd(x))>0.001: #定义精度为0.001
  x=x+a*fd(x)
 if x>=0 and x<3.14*4:
  print(x,f(x))
  dic[y]=x
print(dic)
ls=[]
for i in dic.keys():
 cor=0
 if ls is None:
  ls.append(dic[i])
 else:
  for j in ls:
   if dic[i]-j<0.1:
    cor=1
    break
  if cor==0:
   ls.append(dic[i])
print(ls)