打卡2

导入数据

#首先导入各种库
import warnings
warnings.filterwarnings('ignore')
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
Train_data = pd.read_csv('used_car_train_20200313.csv', sep=' ')
Test_data = pd.read_csv('used_car_testA_20200313.csv', sep=' ')

观察训练集和预测集的大致情况

print(Train_data.head().append(Train_data.tail()))
print('Train_data.shape:',Train_data.shape)
print(Test_data.head().append(Test_data.tail()))
print('Test_data.shape:',Test_data.shape)
print(Train_data.describe())
print(Test_data.describe())
Train_data.info()
Test_data.info()

缺失值处理

# nan可视化
missing = Train_data.isnull().sum()
missing = missing[missing > 0]
missing.sort_values(inplace=True)
missing.plot.bar()
plt.show()
## 2) 查看异常值检测
Train_data['notRepairedDamage'].value_counts()
Train_data['notRepairedDamage'].replace('-', np.nan, inplace=True