[人工智能]数据挖掘常用可视化技巧（特征分布图，PCA可视化，热力图，....）

本文就是为方便自己查代码

绘制特征分布图

plt.style.use('seaborn')

features_set=list(df_graph.columns.difference(['source','target','class']))
fig, axes = plt.subplots(ncols=2, nrows=4,figsize=[15,20])
for i, ax, fea in zip(range(10), axes.flat,features_set):
    ax.set_title(fea)
    ax.set_xlabel('values')
    ax.set_ylabel('nums')
    sns.distplot(df_graph[fea].values,
        hist=True,kde=True, ax=ax,color='royalblue')
plt.show()

绘制热力图

%matplotlib inline
complete_features=df_graph.loc[:,df_graph.columns.difference(['source','taregt','class'])]
# 绘制热力图
plt.figure(figsize=(6,6))
sns.heatmap(complete_features.corr(),annot=True)

PCA降维可视化

from sklearn.decomposition import PCA
from mpl_toolkits.mplot3d import Axes3D 

def plot_pca(num,data,label):
    pca=PCA(n_components=num)
    X_pca=pca.fit_transform(data)
    print(pca.components_)
    # 分割数据
    X_failure=np.array([x for i,x in enumerate(X_pca) if label[i]==1])
    X_healthy=np.array([x for i,x in enumerate(X_pca) if label[i]==0])
    
    if num==3:
        fig = plt.figure(figsize=[10,15])
        ax = Axes3D(fig)   
        #ax.legend(loc='best')
        ax.set_zlabel('Z', fontdict={'size': 15, 'color': 'red'})
        ax.set_ylabel('Y', fontdict={'size': 15, 'color': 'red'})
        ax.set_xlabel('X', fontdict={'size': 15, 'color': 'red'})
        ax.scatter(X_failure[:,0], X_failure[:,1], X_failure[:,2])
        ax.scatter(X_healthy[:,0], X_healthy[:,1], X_healthy[:,2])
    elif num==2:
        plt.figure(figsize=[10,10])
        plt.scatter(X_failure[:,0],X_failure[:,1])
        plt.scatter(X_healthy[:,0],X_healthy[:,1])
    else:
        print('i do not want to work.....')

# to visualize the data without the node_path feature
# %matplotlib auto
X_norm,y=sift_features(df_graph,dislike_fea=['node_path'])
plot_pca(num=3,data=X_norm,label=y)

AUC曲线图

# start to paint the roc curve
from sklearn.metrics import roc_curve
from sklearn.metrics import roc_auc_score as AUC
def plot_roc(df):
    # calculate the FPR(false positive) and Recall(real positive)
    FPR,recall,thresholds = roc_curve(df['label'],df['score'])
    # calculate the area of the AUC
    area=AUC(df['label'],df['score'])
    # plot roc curve
    plt.figure()
    plt.plot(FPR,recall,label='ROC curve (area = %0.2f)'%area)
    plt.legend(loc='lower right')
    plt.show()

plot_roc(Jaccard_df)

有空更新