利用Python对银行进行数据分析

# 导入数据源
import pandas as pd
data = pd.read_csv(r'/Users/faye/Desktop/loan.csv')

• 判断收入与坏账率的关系

# 判断数据是否有缺失
data.info()# 填充月收入缺失值
data = data.fillna({'月收入':data['月收入'].mean()})  # 用均值填充
# 连续值离散化
cut_bins = [0,5000,10000,15000,20000,100000]
income_cut = pd.cut(data['月收入'],cut_bins)
all_income_user = data['好坏客户'].groupby(income_cut).count()
bad_income_user = data['好坏客户'].groupby(income_cut).sum()
bad_rate = bad_income_user / all_income_user
bad_rate
# 绘制月收入与坏账率关系图
bad_rate.plot.bar()

• 判断年龄与坏账率的关系

age_cut = pd.cut(data['年龄'],6)
all_age_user = data['好坏客户'].groupby(age_cut).count()
bad_age_user = data['好坏客户'].groupby(age_cut).sum()
bad_rate = bad_age_user / all_age_user
bad_rate.plot.bar()

• 判断家庭人口数量与坏账率的关系

all_age_user = data.groupby('家属数量')['好坏客户'].count()
bad_age_user = data.groupby('家属数量')['好坏客户'].sum()
bad_rate = bad_age_user / all_age_user
bad_rate.plot()