import numpy as np

from scipy import io

import matplotlib.pyplot as plt

from sklearn import svm

def SVM():

data = np.loadtxt('data.txt', delimiter=',')

X = data[:, :-1]

y = data[:, -1]

plt1 = plot_data(X, y)

plt1.show()

model = svm.SVC(gamma=20).fit(X, y)

plot_decisionBoundary(X, y, model, 'no')

# 线性

data1 = io.loadmat('data1.mat')

X = data1['X']

y = data1['y']

plt1 = plot_data(X, y)

plt1.show()

model = svm.SVC(kernel='linear').fit(X, y)

plot_decisionBoundary(X, y, model)

data2 = io.loadmat('data2.mat')

X = data2['X']

y = data2['y']

plt1 = plot_data(X, y)

plt1.show()

model = svm.SVC(gamma=100).fit(X, y)

plot_decisionBoundary(X, y, model, 'no')

data3 = io.loadmat('data3.mat')

X = data3['X']

y = data3['y']

plt1 = plot_data(X, y)

plt1.show()

model = svm.SVC(gamma=100).fit(X, y)

plot_decisionBoundary(X, y, model, 'no')

# 画决策边界

def plot_decisionBoundary(X, y, model, class_='linear'):

plot = plot_data(X, y)

if class_ == 'linear':

w = model.coef_[0]

b = model.intercept_

xd = np.linspace(np.min(X[:, 0]), np.max(X[:, 1]), 100)

# w0*x + w1*y + b = 0

yd = -(w[0] * xd + b) / w[1]

plot.plot(xd, yd, 'b-')

plot.show()

else:

x1 = np.linspace(np.min(X[:, 0]), np.max(X[:, 0]), 100)

x2 = np.linspace(np.min(X[:, 1]), np.max(X[:, 1]), 100)

X1, X2 = np.meshgrid(x1, x2)

vel = np.zeros(X1.shape)

for i in range(X1.shape[0]):

X = np.hstack((X1[:, i].reshape(-1, 1), X2[:, i].reshape(-1, 1)))

vel[:, i] = model.predict(X)

plot.contour(X1, X2, vel, [0, 1], color='blue')

plot.show()

# 作图

def plot_data(X, y):

y1 = np.where(y == 1)

y0 = np.where(y == 0)

plt.plot(X[y0, 0], X[y0, 1], 'ro', ms=4)

plt.plot(X[y1, 0], X[y1, 1], '^g', ms=4)

return plt

if __name__ == "__main__":

SVM()