-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathBP.py
166 lines (143 loc) · 6.83 KB
/
BP.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
#!/usr/bin/python
# coding=utf-8
from numpy import *
from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
import numpy as np
from pylab import *
import time
import xlrd
import sys
reload(sys)
sys.setdefaultencoding('utf-8')
matplotlib.rcParams['font.sans-serif']=['Droid Sans Fallback']
def read_xls_file(filename): #读取训练数据
data = xlrd.open_workbook(filename)
sheet1 = data.sheet_by_index(0)
m = sheet1.nrows
n = sheet1.ncols
pop = []
veh = []
roa = []
pas = []
fre = []
for i in range(m):
row_data = sheet1.row_values(i)
if i > 0:
pop.append(row_data[1])
veh.append(row_data[2])
roa.append(row_data[3])
pas.append(row_data[4])
fre.append(row_data[5])
dataMat = np.mat([pop,veh,roa])
labels = np.mat([pas,fre])
dataMat_old = dataMat
labels_old = labels
return dataMat,labels,dataMat_old,labels_old
def read_xls_testfile(filename): #读取测试数据
data = xlrd.open_workbook(filename)
sheet1 = data.sheet_by_index(0)
m = sheet1.nrows
n = sheet1.ncols
pop = []
veh = []
roa = []
for i in range(m):
row_data = sheet1.row_values(i)
if i > 0:
pop.append(row_data[1])
veh.append(row_data[2])
roa.append(row_data[3])
dataMat = np.mat([pop,veh,roa])
return dataMat
def Norm(dataMat,labels): #归一化数据
dataMat_minmax = np.array([dataMat.min(axis=1).T.tolist()[0],dataMat.max(axis=1).T.tolist()[0]]).transpose()
dataMat_Norm = ((np.array(dataMat.T)-dataMat_minmax.transpose()[0])/(dataMat_minmax.transpose()[1]-dataMat_minmax.transpose()[0])).transpose()
labels_minmax = np.array([labels.min(axis=1).T.tolist()[0],labels.max(axis=1).T.tolist()[0]]).transpose()
labels_Norm = ((np.array(labels.T).astype(float)-labels_minmax.transpose()[0])/(labels_minmax.transpose()[1]-labels_minmax.transpose()[0])).transpose()
return dataMat_Norm,labels_Norm,dataMat_minmax,labels_minmax
def f(x): #激励函数
return 1/(1+np.exp(-x))
def BP(sampleinnorm, sampleoutnorm): #BP算法
maxepochs = 60000
learnrate = 0.030
errorfinal = 0.65*10**(-3)
samnum = 20
indim = 3
outdim = 2
hiddenunitnum = 3
n,m = shape(sampleinnorm)
w1 = 0.5*np.random.rand(hiddenunitnum,indim)-0.1
b1 = 0.5*np.random.rand(hiddenunitnum,1)-0.1
w2 = 0.5*np.random.rand(outdim,hiddenunitnum)-0.1
b2 = 0.5*np.random.rand(outdim,1)-0.1
errhistory = []
for i in range(maxepochs):
hiddenout = f((np.dot(w1,sampleinnorm).transpose()+b1.transpose())).transpose()
networkout = (np.dot(w2,hiddenout).transpose()+b2.transpose()).transpose()
err = sampleoutnorm - networkout
sse = sum(sum(err**2))/m
errhistory.append(sse)
#if sse < errorfinal:
# break
delta2 = err
delta1 = np.dot(w2.transpose(),delta2)*hiddenout*(1-hiddenout)
dw2 = np.dot(delta2,hiddenout.transpose())
db2 = np.dot(delta2,np.ones((samnum,1)))
dw1 = np.dot(delta1,sampleinnorm.transpose())
db1 = np.dot(delta1,np.ones((samnum,1)))
w2 += learnrate*dw2
b2 += learnrate*db2
w1 += learnrate*dw1
b1 += learnrate*db1
return errhistory,b1,b2,w1,w2,maxepochs
def show(sampleinnorm,sampleoutminmax,sampleout,errhistory,maxepochs): #图形显示
hiddenout = f((np.dot(w1,sampleinnorm).transpose()+b1.transpose())).transpose()
networkout = (np.dot(w2,hiddenout).transpose()+b2.transpose()).transpose()
diff = sampleoutminmax[:,1]-sampleoutminmax[:,0]
networkout2 = networkout
networkout2[0] = networkout2[0]*diff[0]+sampleoutminmax[0][0]
networkout2[1] = networkout2[1]*diff[1]+sampleoutminmax[1][0]
sampleout = np.array(sampleout)
fig,axes = plt.subplots(nrows=2,ncols=1,figsize=(12,10))
line1, = axes[0].plot(networkout2[0],'k',markeredgecolor='b',marker = 'o',markersize=9)
line2, = axes[0].plot(sampleout[0],'r',markeredgecolor='g',marker = u'$\star$',markersize=9)
line3, = axes[0].plot(networkout2[1],'g',markeredgecolor='g',marker = 'o',markersize=9)
line4, = axes[0].plot(sampleout[1],'y',markeredgecolor='b',marker = u'$\star$',markersize=9)
axes[0].legend((line1,line2,line3,line4),(u'客运量预测输出',u'客运量真实输出',u'货运量预测输出',u'货运量真实输出'),loc = 'upper left')
axes[0].set_ylabel(u'公路客运量及货运量')
xticks = range(0,22,1)
xtickslabel = range(1990,2012,1)
axes[0].set_xticks(xticks)
axes[0].set_xticklabels(xtickslabel)
axes[0].set_xlabel(u'年份')
axes[0].set_title(u'BP神经网络')
errhistory10 = np.log10(errhistory)
minerr = min(errhistory10)
plt.plot(errhistory10)
axes[1]=plt.gca()
axes[1].set_yticks([-2,-1,0,1,2,minerr])
axes[1].set_yticklabels([u'$10^{-2}$',u'$10^{-1}$',u'$1$',u'$10^{1}$',u'$10^{2}$',str(('%.4f'%np.power(10,minerr)))])
axes[1].set_xlabel(u'训练次数')
axes[1].set_ylabel(u'误差')
axes[1].set_title(u'误差曲线')
plt.show()
plt.close()
return diff, sampleoutminmax
def pre(dataMat,dataMat_minmax,diff,sampleoutminmax,w1,b1,w2,b2): #数值预测
dataMat_test = ((np.array(dataMat.T)-dataMat_minmax.transpose()[0])/(dataMat_minmax.transpose()[1]-dataMat_minmax.transpose()[0])).transpose()
hiddenout = f((np.dot(w1,dataMat_test).transpose()+b1.transpose())).transpose()
networkout1 = (np.dot(w2,hiddenout).transpose()+b2.transpose()).transpose()
networkout = networkout1
networkout[0] = networkout[0]*diff[0] + sampleoutminmax[0][0]
networkout[1] = networkout[1]*diff[1] + sampleoutminmax[1][0]
print "2010年预测的公路客运量为：", int(networkout[0][0]),"(万人)"
print "2010年预测的公路货运量为：", int(networkout[1][0]),"(万吨)"
print "2011年预测的公路客运量为：", int(networkout[0][1]),"(万人)"
print "2011年预测的公路货运量为：", int(networkout[1][1]),"(万吨)"
dataMat,labels,dataMat_old,labels_old = read_xls_file('ex5data.xlsx')
dataMat_Norm,labels_Norm, dataMat_minmax, labels_minmax = Norm(dataMat,labels)
err, b1, b2, w1, w2,maxepochs = BP(dataMat_Norm,labels_Norm)
dataMat_test = read_xls_testfile('ex5data_test.xlsx')
diff, sampleoutminmax = show(dataMat_Norm,labels_minmax,labels,err,maxepochs)
pre(dataMat_test,dataMat_minmax,diff, sampleoutminmax ,w1,b1,w2,b2)