爱数课实验 | 第七期-基于随机森林的金融危机分析

# 导入相应模块
import numpy as np 
import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.font_manager import FontProperties
# 设置字体
font = FontProperties(fname = "./dataset/SimHei.ttf", size=14)

import seaborn as sns
import random

# 设置绘图风格
%matplotlib inline
sns.set(style='whitegrid')

# 忽略所有警告
import warnings
warnings.filterwarnings('ignore')

# 读取数据
data = pd.read_csv('./dataset/african_crises.csv')
data.sample(5)

unique_countries = data.country.unique()
unique_countries

# 数据集的基本信息
data.______()

# 查看数据统计性指标
data.____________(include = 'all')

#查看货币危机currency_crises的取值为2的数据
data[data['currency_crises'] == 2] 

data = data[data['currency_crises'] != 2]# 得到生成删除货币危机currency_crises的取值为2的数据集
data.______ # 查看新生成的数据集大小

plt.figure(figsize=(12,20))

for i in range(13):
    
    plt.subplot(7,2,i+1)
    country = unique_countries[i]
    
    # 随机生成一种颜色 random.choice():从一个序列中随机的抽取一个元素，抽取6次组成6位代表随机颜色
    col="#"+''.join([random.choice('0123456789ABCDEF') for j in range(6)])
    
    # 绘制折线图
    sns.____________(data[data.country == country]['year'],data[data.country == country]['exch_usd'],label = country,color = col)
    
    # np.logical_and()逻辑与 两个条件均成立时返回True
    plt.plot([np.min(data[np.logical_and(data.country == country,data.independence == 1)]['year']),
              np.min(data[np.logical_and(data.country == country,data.independence == 1)]['year'])],
             [0,np.max(data[data.country == country]['exch_usd'])],color = 'black',linestyle = 'dotted',alpha = 0.8)
    

    plt.______(country) # 添加图像标题
    
plt.tight_layout() # 自动调整子图参数以提供指定的填充
plt.show() # 输出13个国家的货币兑美元的汇率变化情况的折线图

plt.figure(figsize=(12,20))

for i in range(13):
    
    plt.subplot(7,2,i+1)
    country = unique_countries[i]
    
    # 随机生成一种颜色
    col="#"+''.join([random.choice('0123456789ABCDEF') for j in range(6)])
    
    # 绘制折线图
    sns.lineplot(data[data.country == country]['year'],data[data.country == country]['inflation_annual_cpi'],label = country,color = col)
    
    # 加入散点图
    plt.______(data[data.country == country]['year'],data[data.country == country]['inflation_annual_cpi'],color = col,s = 28) # s指散点的面积
    
    plt.plot([np.min(data[np.logical_and(data.country == country,data.independence==1)]['year']),
              np.min(data[np.logical_and(data.country == country,data.independence==1)]['year'])],
             [np.min(data[data.country == country]['inflation_annual_cpi']),np.max(data[data.country == country]['inflation_annual_cpi'])],
             color = 'black',linestyle = 'dotted',alpha = 0.8) # alpha指颜色透明度

    plt.title(country)
    
plt.tight_layout() # 自动调整子图参数以提供指定的填充
plt.show() # 输出13个国家的通货膨胀率变化情况

sns.set(style='darkgrid')
columns = ['systemic_crisis','domestic_debt_in_default','sovereign_external_debt_default','currency_crises','inflation_crises','banking_crisis']

# 绘制其他特征的分布规律图
plt.figure(figsize=(16,16))

for i in range(6):
    plt.subplot('32'+str(i+1))
    sns.countplot(y = data.country,hue = data[columns[i]],palette = 'rocket') # palette为调色板
    plt.______(loc = 0) # 选择最优的图例位置
    plt.title(columns[i])
    
plt.tight_layout()
plt.show()

# 创建包含年份，国家，系统性危机，银行危机的数据集
systemic = data[['year','country', 'systemic_crisis', 'banking_crisis']]

# 绘制观察系统性危机与银行危机发生的重叠性
systemic = systemic[(systemic['country'] == 'Central African Republic') | (systemic['country']=='Kenya') | (systemic['country']=='Zimbabwe') ]
plt.figure(figsize=(12,12))
count = 1

for country in systemic.country.unique():
    plt.subplot(len(systemic.country.unique()),1,count)
    subset = systemic[(systemic['country'] == country)]
    sns.lineplot(subset['year'],subset['systemic_crisis'],ci=None) # ci参数可用于指定线段区间的大小
    plt.scatter(subset['year'],subset["banking_crisis"], color='coral', label='Banking Crisis')
    plt.subplots_adjust(hspace=0.6) # hspace用来设置子图上下间的距离
    plt.______('Years') # 给x轴命名
    plt.______('Systemic Crisis/Banking Crisis') # 给y轴命名
    plt.title(country)
    count+=1

# 将银行危机banking_crisis列进行特征编码
# 将银行危机banking_crisis中未发生危机的数据标为0，发生危机的数据标为1
data['banking_crisis'] = data['banking_crisis'].map({"no_crisis":0,"crisis":1})

# 选出所有特征
selected_features = ['systemic_crisis', 'exch_usd', 'domestic_debt_in_default','sovereign_external_debt_default', 'gdp_weighted_default',
       'inflation_annual_cpi', 'independence', 'currency_crises','inflation_crises','banking_crisis']

corr = data[selected_features].______() # 得到各特征间的相关性大小生成相关性矩阵

fig = plt.figure(figsize = (12,8))

cmap = sns.diverging_palette(220, 10, as_cmap=True) # 生成蓝-白-红的颜色列表
mask = np.zeros_like(corr, dtype=np.bool) # 返回与相关性矩阵具有相同形状和类型的零数组作为掩码
mask[np.triu_indices_from(mask)] = True # 给相关性矩阵的上三角阵生成掩码

# 绘制热力图
sns.______(corr, mask=mask, cmap=cmap,vmin=-0.5,vmax=0.7, center=0,annot = True,
            square=True, linewidths=.5,cbar_kws={"shrink": .5});

plt.title("特征间的相关性",fontproperties = font)
plt.show()

data.drop(['case','cc3'],axis = 1,inplace = True) # 在原数据集上删掉case列和cc3列
data.head()

# 对国家country进行labelencoder
from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()
le.______(data['country'].values) # 将country的值塞入空字典
data['country']=le.____________(data['country'].values) # 将字典中的country的值转变为索引值

print(data['country']) # 查看特征编码后的country名称

# 绘制未发生银行危机no_crisis与发生银行危机crisis的柱状图
fig = plt.figure(figsize = (8,6))

data['banking_crisis'].value_counts().plot(kind='______',rot = 360,color = 'lightseagreen')

plt.xticks([0,1],["no_crisis","crisis"])
plt.show()

# 训练集与测试集的划分
from sklearn import model_selection

x = data.drop('banking_crisis',axis = 1) # 将删除banking_crisis列的数据集作为x
y = data['banking_crisis'] # banking_crisis列作为y

x_train,x_test,y_train,y_test = model_selection.__________________(x, y,test_size=0.2,random_state = 33,stratify=y)

from sklearn.ensemble import RandomForestClassifier

# 训练随机森林分类模型
rf = RandomForestClassifier(n_estimators = 100, max_depth = 20,max_features = 10, random_state = 20)
rf.fit(x_train, y_train) 
y_pred = rf.______(x_test) # 对y进行预测

from sklearn.metrics import classification_report,confusion_matrix

print(classification_report(y_test, y_pred)) # 输出模型的预测报告
confusion_matrix = __________________(y_test, y_pred) 
print(confusion_matrix) # 输出混淆矩阵

# 绘制混淆矩阵热力图
fig,ax = plt.subplots(figsize=(8,6)) 
sns._________(confusion_matrix,ax=ax,annot=True,annot_kws={'size':15}, fmt='d',cmap = 'YlGnBu_r')
ax.set_ylabel('真实值',fontproperties = font)
ax.set_xlabel('预测值',fontproperties = font)
ax.set_title('混淆矩阵热力图',fontproperties = font)
plt.show() # 输出混淆矩阵热力图

from sklearn.metrics import roc_auc_score
from sklearn.metrics import roc_curve

roc_auc = ____________(y_test, rf.predict(x_test)) #计算auc的值
fpr, tpr, thresholds = ____________(y_test, rf.predict_proba(x_test)[:,1]) #计算不同阈值下的TPR和FPR

# 绘制ROC曲线
plt.figure(figsize = (8,6))

plt.plot(fpr, tpr, label='Random Forest (area = %0.2f)' % roc_auc)
plt.plot([0, 1], [0, 1],'r--')# 绘制随机猜测线
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.05])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('ROC curve')

plt.legend(loc="lower right") # 图例位置位于右下方
plt.show()

# 对x_train,y_train进行SMOTE过采样
from imblearn.over_sampling import SMOTE
x_train_resampled, y_train_resampled = SMOTE(random_state=4).fit_resample(x_train, y_train)

print(x_train_resampled.shape, y_train_resampled.shape) #查看采样后的数据集大小

# 通过网格搜索选择最优参数
from sklearn.model_selection import GridSearchCV
param_grid = [{
    'n_estimators':[10,20,30,40,50],
    'max_depth':[5,8,10,15,20,25]
}]
grid_search = GridSearchCV(rf, param_grid, scoring = 'recall')

# 输出最佳参数组合以及分数
grid_search.fit(x_train_resampled, y_train_resampled)

print("best params:", grid_search.best_params_)

print("best score:", grid_search.best_score_)

fig = plt.figure(figsize=(16,12))

# 得到随机森林特征重要性评分
rf_importance = rf.__________________
index = data.drop(['banking_crisis'], axis=1).columns # 删掉银行危机banking_crisis列特征

# 对得到的特征重要性评分进行降序排序
rf_feature_importance = pd.DataFrame(rf_importance.T, index=index,columns=['score']).sort_values(by='score', ascending=True)

# 水平条形图绘制
rf_feature_importance.plot(kind='______',legend=False,color = 'deepskyblue')

plt.title('随机森林特征重要性',fontproperties = font)

plt.show()