df['Weight'] = df['Weight'].astype(float64)
 df['Height'] = df['Height'].astype(float64)
 # 对体重判定结果进行数值化处理
 le = preprocessing.LabelEncoder()
 df['Class_2'] = le.fit_transform(df['Class'])
 df.head()

X = df[['Height', 'Weight']]
 Y = df[['Class_2']]
 x_train, x_test,y_train, y_test=train_test_split(X,Y,train_size=0.7,test_size=0.3)

 params = {'kernel':'linear'}
 classifier = SVC(**params)
 classifier.fit(x_train, y_train.values.ravel())

x_train, x_test,y_train, y_test=train_test_split(X,Y,train_size=0.7,test_size=0.3)
 classifier.fit(x_train, y_train.values.ravel())

 y_min,  y_max = df[['Weight']].values.min()  - 1.0,  df[['Weight']].values.max() + 1.0
 step_size  = 0.1
 x_values,  y_values = np.meshgrid(np.arange(x_min,x_max,step_size),
                                  np.arange(y_min,y_max,step_size))
 mesh_output = classifier.predict(np.c_[x_values.ravel(),y_values.ravel()])
 mesh_output = mesh_output.reshape(x_values.shape)
 plt.figure()

# 预测值
 plt.pcolormesh(x_values,y_values,mesh_output,cmap=pltNaN.Paired, alpha=0.5)
 # 原数据
 plt.scatter(df[['Height']],df[[  'Weight']], c=df[['Class_2']],  s=80, edgecolors='black',  linewidths=1,  marker='o')
 plt.title('肥胖判定（SVM线性分类器）')
 plt.xlabel('身高')
 plt.ylabel('体重')
 plt.show()

 from sklearn.metrics  import classification_report
 print(classification_report(y_test,  classifier.predict(x_test)))