绘制时间序列

Question

我试图绘制分钟的股票数据，这些数据可作为熊猫系列使用。股票数据在上午9:30到下午4:00之间可用。当我绘制它，我得到这样的事情：

有没有办法避免在晚间时间插值？

Answer 1

您将需要构建自己的坐标轴，以排除市场关闭期间的情节。这很烦琐。例子如下：

import pandas as pd 
import matplotlib.pyplot as plt 
import calendar 
from matplotlib.ticker import FixedLocator 

# --- let's fake up some data: 
drng = pd.period_range('2015-04-01 00:00', '2015-04-02 23:59', freq='1min') 
df = pd.DataFrame({'data':np.random.randn(len(drng))}, index=drng) 
df['data'] = df.data.cumsum() 
# let's only keep the fake data for when the market is open 
# market opens at 9.30am and closes at 4pm. 
df = df[((df.index.hour >= 10) | 
     ((df.index.hour == 9) & (df.index.minute >= 30))) & 
     (df.index.hour <= 15)] 

# --- we will need to construct our own index and labels for matplotlib 
#  this is fiddly ... and will vary depending on period being plotted 
#  this works for two days of data ... but you will want to vary for 
#  shorter or longer periods ... 
df['year'] = df.index.year 
df['month'] = pd.Series(df.index.month, index=df.index 
    ).apply(lambda x: calendar.month_abbr[x]) 
df['day'] = df.index.day 
df['hour'] = df.index.hour 
df['minute'] = df.index.minute 
df.index = range(len(df)) 

minorticks = df[df['minute'] == 0].index.tolist() # hours 
majorticks = df[df['day'] != df['day'].shift()].index.tolist() # days 

minorlabels = pd.Series(df.loc[minorticks, 'hour'].astype(str)).tolist() 
majorlabels = pd.Series('\n' + df.loc[majorticks, 'day'].astype(str) + ' ' + 
       df.loc[majorticks, 'month'].astype(str) + ' ' + 
       df.loc[majorticks, 'year'].astype(str)).tolist() 

# --- and plot 
(fig, ax) = plt.subplots(figsize=(8, 4)) 
df['data'].plot(ax = ax) 
ax.xaxis.set_major_locator(FixedLocator(majorticks)) 
ax.xaxis.set_minor_locator(FixedLocator(minorticks)) 
ax.set_xticklabels(minorlabels, minor=True) 
ax.set_xticklabels(majorlabels, minor=False) 
ax.set_xlabel('Time and Date') 
ax.set_ylabel('Index') 
fig.suptitle('Fake Market Data - without closed time periods') 
fig.tight_layout(pad=2) 
plt.show()