二维样本数据和标签回归及LDA代码实现

二维样本数据和标签回归及LDA代码实现

1、先上效果图：

纠正一下横纵坐标，横坐标为x1,纵坐标为x2，这是样本与标签回归的二维结果图。

2、LDA结果图：

图中的是投影直线和分类面

3、上代码：
样本通过高斯分布生成的随机数，标签为+1，-1.

import numpy as np
import matplotlib.pyplot as  plt
from matplotlib.pyplot import MultipleLocator
from sklearn import model_selection
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
from sklearn import metrics

def get_train_data(data_size=100):
    data_label=np.zeros((2*data_size,1))
    #class 1
    x1=np.reshape(np.random.normal(1,0.3,data_size),(data_size,1))
    y1=np.reshape(np.random.normal(1,0.5,data_size),(data_size,1))
    data_train=np.concatenate((x1,y1),axis=1)
    data_label[0:data_size,:]=1
    #class 2
    x2=np.reshape(np.random.normal(-1,0.3,data_size),(data_size,1))
    y2=np.reshape(np.random.normal(-1,0.5,data_size),(data_size,1))
    data_train=np.concatenate((data_train,np.concatenate((x2,y2),axis=1)),axis=0)
    data_label[data_size:2*data_size,:]=-1
    
    return data_train,data_label

def get_test_data(data_size=10):
    testdata_label=np.zeros((2*data_size,1))
    #class 1
    x1=np.reshape(np.random.normal(1,0.3,data_size),(data_size,1))
    y1=np.reshape(np.random.normal(1,0.5,data_size),(data_size,1))
    data_test=np.concatenate((x1,y1),axis=1)
    testdata_label[0:data_size,:]=1
    #class 2
    x2=np.reshape(np.random.normal(-1,0.3,data_size),(data_size,1))
    y2=np.reshape(np.random.normal(-1,0.5,data_size),(data_size,1))
    data_test=np.concatenate((data_test,np.concatenate((x2,y2),axis=1)),axis=0)
    testdata_label[data_size:2*data_size,:]=-1
    
    return data_test,testdata_label

#linear regression
def LR(data,label):
    datalen=len(data) 
    y=label
    ones=np.ones((trainlen,1))
    x=np.concatenate((data,ones),axis=1)
    if np.linalg.det(np.dot(x.T,x)) == 0:
        print('the matrix is singular, cannot do inverse')
    else:
        w=np.dot(np.dot(np.linalg.inv(np.dot(x.T,x)),x.T),y)
        
    return w

#square error
def ressError(y, yHat):
    n=len(y)
    sum1=np.dot((y-yHat).T,(y-yHat))
    error=sum1/n
    return error
#LDA
def LDA(data,label):
    length=len(data)
    x1=data[:int(length/2),:]
    x2=data[int(length/2):,:]
    mu1=np.mean(x1,axis=0)
    mu2=np.mean(x2,axis=0)
    sw=np.dot((x1-mu1).T,(x1-mu1))+np.dot((x2-mu2).T,(x2-mu2))
    w=np.dot(np.linalg.inv(sw),(mu1-mu2))
    
    return w
##########main##########
data_train,data_label=get_train_data()
data_test,testdata_label=get_test_data()
trainlen=len(data_train)
testlen=len(data_test)

##########LDA#########
#classification surface
lda_model = LinearDiscriminantAnalysis(solver='lsqr', shrinkage=None).fit(data_train, data_label)
label_pred = lda_model.predict(data_test)
print(metrics.confusion_matrix(testdata_label, label_pred))
print(metrics.classification_report(testdata_label, label_pred))
h = 0.01
x0, x1 = np.meshgrid(np.arange(-3, 3, h),np.arange(-3, 3, h))
z = lda_model.predict(np.c_[x0.ravel(), x1.ravel()]) 
z = z.reshape(x0.shape)
plt.contourf(x0, x1, z)

#projection line
w=LDA(data_train,data_label)
ax1=plt.subplot(1,1,1)
ax1.scatter(data_train[:int(trainlen/2),0],data_train[:int(trainlen/2),1],s=15,c='r',label='class1')
ax1.scatter(data_train[int(trainlen/2):trainlen,0],data_train[int(trainlen/2):trainlen,1],s=15,c='y',label='class2')
ax1.scatter(data_test[:int(testlen/2),0],data_test[:int(testlen/2),1],c='b',marker='*',label='test_data')
ax1.scatter(data_test[int(testlen/2):testlen,0],data_test[int(testlen/2):testlen,1],c='b',marker='*')
xmin = -data_train[:, 0].max()
ymin = ( w[1] / w[0] ) * xmin
xmax= data_train[:, 0].max()
ymax = ( w[1] / w[0] ) * xmax

#set LAD axis parameters
ax1.set_xlabel('x1')
ax1.set_ylabel('x2')
x_major_locator=MultipleLocator(0.5)
y_major_locator=MultipleLocator(0.5)
plt.plot([xmin, xmax], [ymin, ymax],label='project line')
plt.legend(loc='upper left')
plt.show()

#draw train_data
ax=plt.subplot(1,1,1)
ax.scatter(data_train[:int(trainlen/2),0],data_train[:int(trainlen/2),1],s=15,c='r',label='class1')
ax.scatter(data_train[int(trainlen/2):trainlen,0],data_train[int(trainlen/2):trainlen,1],s=15,c='y',label='class2')

#draw test_data
ax.scatter(data_test[:int(testlen/2),0],data_test[:int(testlen/2),1],c='b',marker='*',label='test_data')
ax.scatter(data_test[int(testlen/2):testlen,0],data_test[int(testlen/2):testlen,1],c='b',marker='*')

# LR:train
what=LR(data_train,data_label)
# yhat=(what[1]/what[0])*data_train[:,0]
yhat=-(what[0]*data_train[:,0]+what[2])/what[1]

#set regress axis parameters 
ax.set_xlabel('x')
ax.set_ylabel('y')
x_major_locator=MultipleLocator(0.5)
y_major_locator=MultipleLocator(0.5)
plt.legend(loc='upper left')
ax.plot(data_train[:,0],yhat,c='g')


#test error
teones=np.ones((testlen,1))
data_test1=np.concatenate((data_test,teones),axis=1)
yHat = np.dot(data_test1,what)
print("square error:",ressError(testdata_label, yHat))

plt.show()

如果有帮助就留下个赞吧！AOA