##注入所需库
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import random
import numpy as np
import time
import shap
from sklearn.svm import SVC #支持向量机分类器
# from sklearn.neighbors import KNeighborsClassifier #K近邻分类器
# from sklearn.linear_model import LogisticRegression #逻辑回归分类器
import xgboost as xgb #XGBoost分类器
import lightgbm as lgb #LightGBM分类器
from sklearn.ensemble import RandomForestClassifier #随机森林分类器
# from catboost import CatBoostClassifier #CatBoost分类器
# from sklearn.tree import DecisionTreeClassifier #决策树分类器
# from sklearn.naive_bayes import GaussianNB #高斯朴素贝叶斯分类器
from skopt import BayesSearchCV
from skopt.space import Integer
from deap import base, creator, tools, algorithms
from sklearn.model_selection import StratifiedKFold, cross_validate # 引入分层 K 折和交叉验证工具
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score # 用于评估分类器性能的指标
from sklearn.metrics import classification_report, confusion_matrix #用于生成分类报告和混淆矩阵
from sklearn.metrics import make_scorer#定义函数
import warnings #用于忽略警告信息
warnings.filterwarnings("ignore") # 忽略所有警告信息
#聚类
from sklearn.cluster import KMeans, DBSCAN, AgglomerativeClustering
from sklearn.preprocessing import StandardScaler
from sklearn.decomposition import PCA
from sklearn.metrics import silhouette_score, calinski_harabasz_score, davies_bouldin_score
#3D可视化
from mpl_toolkits.mplot3d import Axes3D
#设置中文字体&负号正确显示
plt.rcParams['font.sans-serif']=['STHeiti']
plt.rcParams['axes.unicode_minus']=True
plt.rcParams['figure.dpi']=100
#读取数据&查看数据基本信息
data=pd.read_csv(r'data.csv')
print(f'{data.info()}\n{data.isnull().sum()}')
#绘制图像
# plt.figure(figsize=(6,4))
# sns.boxenplot(x=data['Annual Income'])
# plt.title('年收入箱线图')
# plt.tight_layout()
# plt.show()
# plt.figure(figsize=(6,4))
# sns.boxenplot(
# x='Credit Default',
# y='Monthly Debt',
# data=data.dropna()
# )
# plt.title('月欠款分类箱线图')
# plt.xticks([0,1],['n','y'])
# plt.tight_layout()
# plt.show()
# plt.figure(figsize=(6,4))
# sns.histplot(
# x='Monthly Debt',
# hue='Credit Default',
# hue_order=[0,1],
# data=data.dropna(),
# kde=True,
# element='bars'
# )
# plt.title('月欠款分类直方图')
# plt.xlabel('Credit Default')
# plt.ylabel('count')
# plt.legend(labels=['n','y'])
# plt.tight_layout()
# plt.show()
features=['Annual Income','Months since last delinquent','Number of Credit Problems','Monthly Debt']
# #绘制箱线图子图
# fig,axes=plt.subplots(2,2,figsize=(6,4))
# for i,feature in enumerate(features):
# row,col=i//2,i%2
# sns.boxenplot(x=data[feature].dropna(),ax=axes[row,col])
# axes[row,col].set_title(f'boxplot of {feature}')
# axes[row,col].set_xlabel(feature)
# plt.tight_layout()
# plt.show()
# #绘制分类箱线图子图
# fig,axes=plt.subplots(2,2,figsize=(6,4))
# for i,feature in enumerate(features):
# row,col=i//2,i%2
# sns.boxenplot(
# x='Credit Default',
# y=feature,
# data=data.dropna(),
# ax=axes[row,col]
# )
# axes[row,col].set_title(f'boxplot of {feature}')
# axes[row,col].set_label(feature)
# axes[row,col].set_xticks([0,1],['n','y'])
# plt.tight_layout()
# plt.show()
# #绘制分类箱线直方图
# fig,axes=plt.subplots(2,2,figsize=(6,4))
# for i,feature in enumerate(features):
# row,col=i//2,i%2
# sns.histplot(
# x=feature,
# hue='Credit Default',
# hue_order=(0,1),
# data=data.dropna(),
# kde=True,
# element='bars',
# ax=axes[row,col]
# )
# axes[row,col].set_title(f'histplot of {feature}')
# axes[row,col].set_xlabel(feature)
# axes[row,col].set_ylabel('count')
# axes[row,col].legend(labels=['n','y'])
# plt.tight_layout()
# plt.show()
#数据填补
for i in data.columns:
if data[i].dtype!='object':
if data[i].isnull().sum()>0:
data[i].fillna(data[i].mean(),inplace=True)
else:
if data[i].isnull().sum()>0:
data[i].fillna(data[i].mode()[0],inplace=True)
mapping={'10+ years':0,
'9 years':1,
'8 years':2,
'7 years':3,
'6 years':4,
'5 years':5,
'4 years':6,
'3 years':7,
'2 years':8,
'1 year':9,
'< 1 year':10}
data['Years in current job']=data['Years in current job'].map(mapping)
dummies_list=[]
data2=pd.read_csv(r'data.csv')
data=pd.get_dummies(data=data,drop_first=True)
for i in data.columns:
if i not in data2.columns:
dummies_list.append(i)
for i in dummies_list:
data[i]=data[i].astype(int)
print(f'{data.info()}\n{data.columns}')
# #绘制相关热力图
# continuous_features=['Annual Income', 'Years in current job', 'Tax Liens',
# 'Number of Open Accounts', 'Years of Credit History',
# 'Maximum Open Credit', 'Number of Credit Problems',
# 'Months since last delinquent', 'Bankruptcies', 'Current Loan Amount',
# 'Current Credit Balance', 'Monthly Debt', 'Credit Score']
# confusion_martix=data[continuous_features].corr()
# plt.figure(figsize=(12,10))
# sns.heatmap(confusion_martix,annot=True,cmap='coolwarm',vmin=-1,vmax=1)
# plt.title('相关热力图')
# plt.xticks(rotation=30,ha='right')
# plt.tight_layout()
# plt.show()
#划分数据集
from sklearn.model_selection import train_test_split
x=data.drop('Credit Default',axis=1)
y=data['Credit Default']
x_train,x_test,y_train,y_test=train_test_split(x,y,test_size=0.2,random_state=42)
print(f'train:{x_train.shape}\ntest:{x_test.shape}')
# #模型训练
# #Randomforrest
# start_time=time.time()
# rf_model=RandomForestClassifier(random_state=42,class_weight='balanced')
# rf_model.fit(x_train,y_train)
# rf_pred=rf_model.predict(x_test)
# end_time=time.time()
# print(f'训练与预测耗时:{end_time-start_time:.4f}')
# print('\n随机森林分类报告')
# print(classification_report(y_test,rf_pred))
# print('\n随机森林混淆矩阵')
# print(confusion_matrix(y_test,rf_pred))
#定义约登系数
def youden_score(y_true,y_pred):
tn,fp,fn,tp=confusion_matrix(y_true,y_pred).ravel()
sensitivity=tp/(tp+fn)
specificity=tn/(tn+fp)
return sensitivity+specificity-1
youden_scorer=make_scorer(youden_score)
# #SMOTE过采样后带权重网格搜索优化的随机森林
# #SMOTE
# from imblearn.over_sampling import SMOTE
# smote=SMOTE(random_state=42)
# x_train_smote,y_train_smote=smote.fit_resample(x_train,y_train)
# #网格搜索&交叉验证
# from sklearn.model_selection import GridSearchCV
# cv_strategy=StratifiedKFold(n_splits=5,shuffle=True,random_state=42)
# param_grid={
# 'n_estimators':[5,10,15],
# 'max_depth':[None,5,10],
# 'min_samples_split':[2,3,4],
# 'min_samples_leaf':[2,3,4]
# }
# grid_search=GridSearchCV(
# estimator=RandomForestClassifier(random_state=42,class_weight='balanced'),
# param_grid=param_grid,
# cv=cv_strategy,
# n_jobs=-1,
# scoring=youden_scorer
# )
# start_time=time.time()
# grid_search.fit(x_train_smote,y_train_smote)
# end_time=time.time()
# best_model=grid_search.best_estimator_
# best_pred=best_model.predict(x_test)
# print(f'网格搜索耗时:{end_time-start_time:.4f}秒')
# print('最佳参数:',grid_search.best_params_)
# print('\n带权重网格搜索优化后的随机森林在测试集上的分类报告')
# print(classification_report(y_test,best_pred))
# print('网格搜索优化后的随机森林在测试集上的混淆矩阵')
# print(confusion_matrix(y_test,best_pred))
# #SMOTE过采样后带权重的贝叶斯优化的随机森林
# #SMOTE过采样
# from imblearn.over_sampling import SMOTE
# smote=SMOTE(random_state=42)
# x_train_smote,y_train_smote=smote.fit_resample(x_train,y_train)
# #贝叶斯优化&交叉验证
# from sklearn.model_selection import GridSearchCV
# cv_strategy=StratifiedKFold(n_splits=5,shuffle=True,random_state=42)
# search_space={
# 'n_estimators':Integer(1,5),
# 'max_depth':Integer(1,5),
# 'min_samples_split':Integer(2,6),
# 'min_samples_leaf':Integer(1,5)
# }
# bayes_search=BayesSearchCV(
# estimator=RandomForestClassifier(random_state=42,class_weight='balanced'),
# search_spaces=search_space,
# cv=cv_strategy,
# n_jobs=-1,
# scoring=youden_scorer
# )
# start_time=time.time()
# bayes_search.fit(x_train_smote,y_train_smote)
# end_time=time.time()
# best_model=bayes_search.best_estimator_
# best_pred=best_model.predict(x_test)
# print(f'贝叶斯优化耗时:{end_time-start_time:.4f}秒')
# print('最佳参数',bayes_search.best_params_)
# print('\n贝叶斯优化后的随机森林在测试集上的分类报告')
# print(classification_report(y_test,best_pred))
# print('\n贝叶斯优化后的随机森林在测试集上侧混淆矩阵')
# print(confusion_matrix(y_test,best_pred))
# #shap分析
# start_time=time.time()
# explainer=shap.TreeExplainer(best_model)
# shap_values=explainer.shap_values(x_test)
# end_time=time.time()
# print(f"shap分析耗时: {end_time - start_time:.4f} 秒")
# print('shap_values shape:',shap_values.shape)
# print('shap_values[0,:,:]shape',shap_values[0,:,:].shape)##这里也可以省略后面的写作shap_values[0】，代表第一个样本所有特征对所有类别的贡献，后面部位可以省略但是前面不能。
# print('shap_values[:,:,0]shape',shap_values[:,:,0].shape)
# print('x_test shape',x_test.shape)
# ##SHAP特征重要性条形图 (Summary Plot - bar)
# print("--- 2. SHAP 特征重要性条形图 ---")
# shap.summary_plot(
# shap_values[:,:,0],
# x_test,
# plot_type='bar',
# show=False
# )
# plt.title('shap特征重要性条形图')
# plt.tight_layout()
# plt.show()
# ##SHAP特征重要性蜂巢图
# print("--- 2. SHAP 特征重要性蜂巢图 ---")
# shap.summary_plot(
# shap_values[:,:,0],
# x_test,
# plot_type='violin',
# show=False
# )
# plt.title(' SHAP特征重要性小提琴图')
# plt.tight_layout()
# plt.show()
#标准化数据，将自变量标准化，聚类就是从自变量中聚合新的自变量，与因变量无关
scaler=StandardScaler()
x_scaled=scaler.fit_transform(x)
#Kmeans++聚类
k_range=range(2,5)
inertia_value=[]
silhouette_scores=[]
ch_scores=[]
db_scores=[]
start_time=time.time()
for k in k_range:
kmeans=KMeans(n_clusters=k,random_state=42)
kmeans_label=kmeans.fit_predict(x_scaled)#提供了每个数据点所属的簇的信息，用于区分不同簇的数据点
inertia_value.append(kmeans.inertia_)
silhouette=silhouette_score(x_scaled,kmeans_label)
silhouette_scores.append(silhouette)
ch=calinski_harabasz_score(x_scaled,kmeans_label)
ch_scores.append(ch)
db=davies_bouldin_score(x_scaled,kmeans_label)
db_scores.append(db)
# print(f'k={k}\n 惯性:{kmeans.inertia_:.2f}\n轮廓系数:{silhouette:.3f}\n CH系数:{ch:.2f}\n DB{db:.3f}')
end_time=time.time()
print(f'聚类分析耗时：{end_time-start_time:.4f}')
# #绘制评估指标图
# plt.figure(figsize=(12,6))
# #肘部法则图
# plt.subplot(2,2,1)
# plt.plot(k_range,inertia_value,marker='o')
# plt.title('肘部法则确定最优聚类数 k(惯性，越小越好)')
# plt.xlabel('聚类数 (k)')
# plt.ylabel('惯性')
# plt.grid(True)
# #轮廓系数图
# plt.subplot(2,2,2)
# plt.plot(k_range,silhouette_scores,marker='o',color='orange')
# plt.title('轮廓系数确定最优聚类数 k(越大越好)')
# plt.xlabel('聚类数 (k)')
# plt.ylabel('轮廓系数')
# plt.grid(True)
# #CH指数图
# plt.subplot(2,2,3)
# plt.plot(k_range,ch_scores,marker='o',color='red')
# plt.title('Calinski-Harabasz 指数确定最优聚类数 k(越大越好)')
# plt.xlabel('聚类数 (k)')
# plt.ylabel('CH 指数')
# plt.grid(True)
# #DB指数图
# plt.subplot(2,2,4)
# plt.plot(k_range,db_scores,marker='o',color='yellow')
# plt.xlabel('聚类数 (k)')
# plt.ylabel('DB 指数')
# plt.grid(True)
# plt.tight_layout()
# plt.show()
# #选择K值进行聚类
# selected_k=20
# kmeans=KMeans(n_clusters=selected_k,random_state=42)
# kmeans_label=kmeans.fit_predict(x_scaled)
x['KMeans_Cluster']=kmeans_label
# #PCA降维
# pca=PCA(n_components=2)
# x_pca=pca.fit_transform(x_scaled)
# #聚类可视化
# plt.figure(figsize=(6,5))
# sns.scatterplot(
# x=x_pca[:,0],
# y=x_pca[:,1],
# hue=kmeans_label,
# palette='viridis'
# )
# plt.title(f'KMean Clustering with k={selected_k} (PCA Visualization)')
# plt.xlabel('PCA Component 1')
# plt.ylabel('PCA Component 2')
# plt.show()
# ##打印KMeans聚类标签前几行
# print(f'KMeans Clueter labels(k={selected_k}added to X):')
# print(x[['KMeans_Cluster']].value_counts())
start_time=time.time()
x1=x.drop('KMeans_Cluster',axis=1)
y1=x['KMeans_Cluster']
rf1_model=RandomForestClassifier(random_state=42,class_weight='balanced')
rf1_model.fit(x1,y1)
explainer=shap.TreeExplainer(rf1_model)
shap_values=explainer.shap_values(x1)
print(shap_values.shape)
# --- 1. SHAP 特征重要性条形图 (Summary Plot - Bar) ---
print("--- 1. SHAP 特征重要性条形图 ---")
shap.summary_plot(shap_values[:,:,0],x1,plot_type='bar',show=False)
plt.title("SHAP Feature Importance (Bar Plot)")
plt.tight_layout()
plt.show()
end_time=time.time()
print(f'SHAP分析耗时:{end_time-start_time:.4f}')
selected_features=['Purpose_debt consolidation','Home Ownership_Home Mortgage','Purpose_home improvements','Purpose_other']
for feature in selected_features:
unique_count=x[feature].nunique()
print(f'{feature}的唯一值数量:{unique_count}')
if unique_count<10:
print(f'{feature}可能是离散型变量')
else:
print(f'{feature}可能是连续性变量')
fig,axes=plt.subplots(2,2,figsize=(10,8))
axes=axes.flatten()
for i,feature in enumerate(selected_features):
axes[i].hist(x[feature],bins=10)
axes[i].set_title(f'histogram of {feature}')
axes[i].set_xlabel(feature)
axes[i].set_ylabel('frequency')
plt.tight_layout()
plt.show()
print(x[['KMeans_Cluster']].value_counts())
x_cluster0=x[x['KMeans_Cluster']==0]
x_cluster1=x[x['KMeans_Cluster']==1]
x_cluster2=x[x['KMeans_Cluster']==2]
x_cluster3=x[x['KMeans_Cluster']==3]
#簇0
fig,axes=plt.subplots(2,2,figsize=(6,4))
axes=axes.flatten()
for i,feature in enumerate(selected_features):
axes[i].hist(x_cluster0[feature],bins=20)
axes[i].set_title(f'Histogram of {feature}')
axes[i].set_xlabel(feature)
axes[i].set_ylabel('frequency')
plt.tight_layout()
plt.show()
#簇1
fig,axes=plt.subplots(2,2,figsize=(6,4))
axes=axes.flatten()
for i,feature in enumerate(selected_features):
axes[i].hist(x_cluster1[feature],bins=20)
axes[i].set_title(f'Histogram of {feature}')
axes[i].set_xlabel(feature)
axes[i].set_ylabel('frequency')
plt.tight_layout()
plt.show()
#簇2
fig,axes=plt.subplots(2,2,figsize=(6,4))
axes=axes.flatten()
for i,feature in enumerate(selected_features):
axes[i].hist(x_cluster2[feature],bins=20)
axes[i].set_title(f'Histogram of {feature}')
axes[i].set_xlabel(feature)
axes[i].set_ylabel('frequency')
plt.tight_layout()
plt.show()
#簇3
fig,axes=plt.subplots(2,2,figsize=(6,4))
axes=axes.flatten()
for i,feature in enumerate(selected_features):
axes[i].hist(x_cluster3[feature],bins=20)
axes[i].set_title(f'Histogram of {feature}')
axes[i].set_xlabel(feature)
axes[i].set_ylabel('frequency')
plt.tight_layout()
plt.show()
print("--- 递归特征消除 (RFE) ---")
from sklearn.feature_selection import RFE
start_time=time.time()
base_model=RandomForestClassifier(random_state=42,class_weight='balanced')
rfe=RFE(base_model,n_features_to_select=5)
rfe.fit(x_train,y_train)
x_train_rfe=rfe.transform(x_train)
x_test_rfe=rfe.transform(x_test)
selected_features_rfe=x_train.columns[rfe.support_].tolist()
print(f"RFE筛选后保留的特征数量: {len(selected_features_rfe)}")
print(f"保留的特征: {selected_features_rfe}")
#训练随机森林模型
rf_model_rfe=RandomForestClassifier(random_state=42,class_weight='balanced')
rf_model_rfe.fit(x_train_rfe,y_train)
rf_pred_rfe=rf_model_rfe.predict(x_test_rfe)
end_time=time.time()
print(f"训练与预测耗时: {end_time - start_time:.4f} 秒")
print("\nRFE筛选后随机森林在测试集上的分类报告:")
print(classification_report(y_test, rf_pred_rfe))
print("RFE筛选后随机森林在测试集上的混淆矩阵:")
print(confusion_matrix(y_test, rf_pred_rfe))
2355
扫一扫
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
