数据分析——pandas时间序列

1. time模块三角转换关系

2. 练习一

统计出911数据中不同月份的电话次数

• 代码

import numpy as np
import pandas as pd

noo_df = pd.read_csv('911.csv')
print(noo_df.info())
print(noo_df.head())

noo_df['timeStamp'] = pd.to_datetime(noo_df['timeStamp'])
noo_df.set_index('timeStamp',inplace=True)
print(noo_df.head())

month_data = noo_df.resample('M')[['title']].count()
print(month_data)

• 运行结果

3. 练习二

北上广深与沈阳5个城市空气质量数据，绘制出北京的PM2.5随时间的变化情况

• 代码

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
plt.rcParams['font.sans-serif'] = ['SimHei'] 
plt.rcParams['axes.unicode_minus'] = False

bj_df = pd.read_csv('PM2.5/BeijingPM20100101_20151231.csv')
print(bj_df.info())
print(bj_df.head())

period_data = pd.PeriodIndex(year=bj_df['year'],month=bj_df['month'],day=bj_df['day'],hour=bj_df['hour'],freq='H')
bj_df['timeStamp'] = period_data
bj_df.set_index('timeStamp',inplace=True)
print(bj_df.head())

resample_data = bj_df.resample('7D')[['PM_US Post']].mean()
print(resample_data)

x = resample_data.index
y = resample_data['PM_US Post'].values

plt.figure(figsize=(10,6))
plt.plot(range(len(x)),y)
plt.xticks(range(len(x))[::10],x[::10],rotation=90)
plt.xlabel('日期')
plt.ylabel('PM2.5平均浓度')
plt.title('北京PM2.5随时间变化情况')
plt.show()

• 运行结果