一、Python所有方向的学习路线
Python所有方向路线就是把Python常用的技术点做整理,形成各个领域的知识点汇总,它的用处就在于,你可以按照上面的知识点去找对应的学习资源,保证自己学得较为全面。
二、学习软件
工欲善其事必先利其器。学习Python常用的开发软件都在这里了,给大家节省了很多时间。
三、入门学习视频
我们在看视频学习的时候,不能光动眼动脑不动手,比较科学的学习方法是在理解之后运用它们,这时候练手项目就很适合了。
网上学习资料一大堆,但如果学到的知识不成体系,遇到问题时只是浅尝辄止,不再深入研究,那么很难做到真正的技术提升。
一个人可以走的很快,但一群人才能走的更远!不论你是正从事IT行业的老鸟或是对IT行业感兴趣的新人,都欢迎加入我们的的圈子(技术交流、学习资源、职场吐槽、大厂内推、面试辅导),让我们一起学习成长!
plt.show()
7.3 结果
最小值点
[[ 0.01025158]
[ 0.05412669]
[ 0.01478472]
[ 0.11712758]
[-0.0201138 ]
[ 0.10935337]
[-0.01210524]
[-0.06925395]
[ 0.06425366]
[ 0.06373066]]
最小值
0.042467753287183725
Process finished with exit code 0
8 案例2
8.1 案例
8.2 Python实现
import numpy as np
import matplotlib.pyplot as plt
import matplotlib as mpl
import matplotlib; matplotlib.use(‘TkAgg’)
mpl.rcParams[‘font.sans-serif’] = [‘SimHei’] # 指定默认字体
mpl.rcParams[‘axes.unicode_minus’] = False # 解决保存图像是负号’-'显示为方块的问题
#=====初始化参数
D=2 #变量维数 ,对应x,y
Xs=1 #搜索区间上限
Xx=-1 #搜索区间下限
#冷却表参数
L = 300 #马可夫链长度 #在温度为t情况下的迭代次数
K = 0.98 #衰减参数
S = 0.01 #步长因子
T=100 #初始温度
YZ = 1e-8 #容差(运行t-1,,t时刻的误差)
P = 0 #Metropolis过程中总接受点
eloss=0.1 #允许的惩罚项误差
#随机选点 初值设定
PreX = np.zeros(shape=(D,1))
PreX[0]=np.random.uniform(1,2,1)#X范围【1,2】
PreX[1]=np.random.uniform(-1,0,1)#y范围【-1,0】
PreBestX = PreX#t-1代的全局最优X
PreX = np.zeros(shape=(D,1))
PreX[0]=np.random.uniform(1,2,1)#X范围【1,2】
PreX[1]=np.random.uniform(-1,0,1)#y范围【-1,0】
BestX = PreX#t时刻的全局最优X
#==目标函数=
def func1(X):
A = 10
pi = np.pi
x = X[0]
y = X[1]
return 2 * A + x ** 2 - A * np.cos(2 * pi * x) + y ** 2 - A * np.cos(2 * pi * y)
#惩罚项=
def calc_e(X):
“”“计算个体的目惩罚项,X 的维度是 size * 2 “””
ee = 0
“”“计算第一个约束的惩罚项”“”
e1 = X[0] + X[1] - 6
ee += max(0, e1)
“”“计算第二个约束的惩罚项”“”
e2 = 3 * X[0] - 2 * X[1] - 5
ee += max(0, e2)
return ee
#==每迭代一次退火一次(降温), 直到满足迭代条件为止=
deta=np.abs(func1(BestX)-func1(PreBestX))#前后能量差
trace=[]#记录
while (deta > YZ) and (T>0.1):#如果能量差大于允许能量差 或者温度大于阈值
T = K * T#降温
print(T)
=在当前温度T下迭代次数=
for i in range(L): #
=在此点附近随机选下一点==
NextX = PreX + S * (np.random.uniform(size=(D, 1)) * (Xs - Xx) + Xx)
===边界条件处理
while NextX[0] > 2 or NextX[0] < 1:#x属于【1,2】
NextX[0] =PreX[0] + S* (np.random.uniform() * (Xs - Xx) + Xx)
while NextX[1] > 0 or NextX[0] < -1: # y属于【-1,0】
NextX[1] = PreX[1] + S* (np.random.uniform() * (Xs - Xx) + Xx)
===是否全局最优解 ===
if (func1(BestX) > func1(NextX)) and (calc_e(BestX)<eloss) and (calc_e(NextX)<eloss):#两者都没有违反约束,下一代适应度更好时选下一代
保留上一个最优解
PreBestX = BestX
此为新的最优解
BestX = NextX
elif (calc_e(BestX)>eloss) and (calc_e(NextX)>eloss) and (func1(BestX) > func1(NextX)):#两者都违反约束,下一代适应度更好时(选下一代)
保留上一个最优解
PreBestX = BestX
此为新的最优解
BestX = NextX
elif (calc_e(BestX)>eloss) and (calc_e(NextX)<eloss) :#上一代惩罚项不满足约束,下一代惩罚项满足约束(选下一代)
保留上一个最优解
PreBestX = BestX
此为新的最优解
BestX = NextX
==Metropolis过程 当代温度T 前后个体判断=
if (func1(PreX) - func1(NextX) > 0) and (calc_e(PreX)<eloss) and(calc_e(NextX)<eloss) : # 两个惩罚项都满足约束,后一个的适应度更小(选后一个)
接受新解
PreX = NextX
P = P + 1
elif (calc_e(PreX)>eloss) and(calc_e(NextX)>eloss) and (func1(PreX) - func1(NextX) > 0) :#两个惩罚项都不满足时,后一个适应度比前一个小(选后一个)
接受新解
PreX = NextX
P = P + 1
elif (calc_e(PreX)>eloss) and(calc_e(NextX)<eloss) :#前一个惩罚项不满足,后一个满足(选后一个)
接受新解
PreX = NextX
P = P + 1
else:
changer = -1 * (func1(NextX) - func1(PreX)) / T
p1 = np.exp(changer)
接受较差的解
if p1 > np.random.random():
PreX = NextX
P = P + 1
trace.append(func1(BestX))
deta = np.abs(func1(BestX) - func1(PreBestX)) # 修改前后能量差
print(‘最优值X\n’,BestX)
print(‘最优目标函数值\n’,func1(BestX))
print(‘最优值对应的惩罚项’,calc_e(BestX))
plt.plot(trace,label=‘迭代曲线’)
plt.xlabel(‘迭代次数’)
plt.ylabel(‘目标函数值’)
plt.show()
8.3 结果
98.0
96.03999999999999
94.11919999999999
92.23681599999999
90.39207968
88.58423808639999
86.812553324672
85.07630225817856
83.37477621301498
81.70728068875468
80.07313507497959
78.47167237347999
76.90223892601038
75.36419414749018
73.85691026454037
72.37977205924956
70.93217661806457
69.51353308570329
68.12326242398922
66.76079717550944
65.42558123199925
64.11706960735926
62.83472821521207
61.57803365090783
60.34647297788967
59.13954351833188
57.95675264796524
56.79761759500593
55.66166524310581
54.54843193824369
53.45746329947881
52.38831403348923
51.34054775281945
50.313736797763056
49.3074620618078
48.32131282057164
47.35488656416021
46.407788832877
45.47963305621946
44.57004039509507
43.67863958719317
42.80506679544931
41.94896545954032
41.10998615034951
40.287786427342525
39.482030698795676
38.69239008481976
37.918542283123365
37.160171437460896
36.41696800871168
35.688628648537446
34.9748560755667
34.27535895405536
33.58985177497426
32.91805473947477
32.25969364468527
31.614499771791564
30.98220977635573
30.362565580828615
29.755314269212043
29.160207983827803
28.577003824151248
28.005463747668223
27.445354472714858
26.89644738326056
26.35851843559535
25.831348066883443
25.314721105545775
24.80842668343486
24.312258149766162
23.826012986770838
23.349492727035422
22.88250287249471
22.42485281504482
21.97635575874392
21.53682864356904
21.10609207069766
20.683970229283705
20.27029082469803
19.86488500820407
19.467587308039988
19.07823556187919
18.696670850641606
18.322737433628774
17.956282684956197
17.597157031257073
17.24521389063193
16.900309612819292
16.562303420562905
16.231057352151648
15.906436205108616
15.588307481006444
15.276541331386316
14.971010504758588
14.671590294663416
14.378158488770147
14.090595318994744
13.808783412614849
13.532607744362553
13.261955589475301
12.996716477685794
12.736782148132079
12.482046505169437
12.232405575066048
11.987757463564726
11.748002314293432
11.513042268007563
11.282781422647412
11.057125794194464
10.835983278310575
10.619263612744364
10.406878340489476
10.198740773679686
9.994765958206093
9.79487063904197
9.598973226261132
9.406993761735908
9.21885388650119
9.034476808771165
8.853787272595742
8.676711527143826
8.50317729660095
8.33311375066893
8.166451475655553
8.003122446142442
7.8430599972195925
7.686198797275201
7.532474821329696
7.381825324903103
7.23418881840504
7.089505042036939
6.9477149411962005
6.808760642372277
6.672585429524831
6.539133720934334
6.408351046515647
6.280184025585334
6.154580345073628
6.031488738172155
5.910858963408712
5.792641784140538
5.676788948457727
5.563253169488572
5.451988106098801
5.342948343976825
5.236089377097288
5.131367589555342
5.028740237764235
4.92816543300895
4.829602124348772
4.733010081861796
4.63834988022456
4.545582882620068
4.454671224967667
4.365577800468314
4.278266244458948
4.192700919569768
4.108846901178373
4.026669963154805
3.9461365638917085
3.8672138326138743
3.7898695559615967
3.714072164842365
3.6397907215455176
3.566994907114607
3.495655008972315
3.425741908792869
3.3572270706170113
3.290082529204671
3.224280878620578
3.159795261048166
3.0965993558272027
3.0346673687106587
2.9739740213364456
2.9144945409097165
2.856204650091522
2.7990805570896913
2.7430989459478976
2.6882369670289394
2.6344722276883608
2.5817827831345936
2.530147127471902
2.4795441849224638
2.4299533012240144
2.381354235199534
2.3337271504955432
2.2870526074856325
2.24131155533592
2.1964853242292017
2.1525556177446177
2.1095045053897254
2.0673144152819307
2.025968126976292
1.985448764436766
1.9457397891480306
1.9068249933650698
1.8686884934977683
1.831314723627813
1.7946884291552565
1.7587946605721514
1.7236187673607084
1.6891463920134941
1.6553634641732242
1.6222561948897598
1.5898110709919646
1.5580148495721253
1.5268545525806827
1.496317461529069
1.4663911122984876
1.4370632900525178
1.4083220242514674
1.3801555837664379
1.352552472091109
1.325501422649287
1.2989913941963012
1.273011566312375
1.2475513349861276
1.222600308286405
1.198148302120677
1.1741853360782633
1.150701629356698
1.127687596769564
1.1051338448341728
1.0830311679374893
1.0613705445787396
1.0401431336871647
1.0193402710134214
0.9989534655931529
0.9789743962812898
0.959394908355664
0.9402070101885507
0.9214028699847796
0.9029748125850839
0.8849153163333823
0.8672170100067146
0.8498726698065803
0.8328752164104487
0.8162177120822397
0.7998933578405949
0.783895490683783
感谢每一个认真阅读我文章的人,看着粉丝一路的上涨和关注,礼尚往来总是要有的:
① 2000多本Python电子书(主流和经典的书籍应该都有了)
② Python标准库资料(最全中文版)
③ 项目源码(四五十个有趣且经典的练手项目及源码)
④ Python基础入门、爬虫、web开发、大数据分析方面的视频(适合小白学习)
⑤ Python学习路线图(告别不入流的学习)
网上学习资料一大堆,但如果学到的知识不成体系,遇到问题时只是浅尝辄止,不再深入研究,那么很难做到真正的技术提升。
一个人可以走的很快,但一群人才能走的更远!不论你是正从事IT行业的老鸟或是对IT行业感兴趣的新人,都欢迎加入我们的的圈子(技术交流、学习资源、职场吐槽、大厂内推、面试辅导),让我们一起学习成长!