使用 matplotlib 绘制两个时间序列数据帧和一个数据帧中的一些缺失日期

Question

我有两个 X 轴上有日期的 DataFrame。我想用两个比例将它们绘制在一个图中。

数据帧 #1

print(df_daily.head())
        date  senti-pol  senti-vader
0 2019-10-01  -0.060639    -0.174223
1 2019-10-02  -0.080265     0.090761
2 2019-10-03  -0.186335    -0.645464
3 2019-10-04   0.014124    -0.043164
4 2019-10-05  -0.035157     0.275379

数据帧 #2

print(df_dbk.head())
             Open   High    Low  Close  Adj Close    Volume
Date                                                       
2019-10-02  6.650  6.720  6.560  6.566      6.566  15527318
2019-10-04  6.520  6.531  6.369  6.480      6.480  16042648
2019-10-07  6.489  6.489  6.348  6.481      6.481  11130966
2019-10-08  6.515  6.529  6.205  6.304      6.304  13736758
2019-10-09  6.300  6.375  6.256  6.294      6.294   8625379

第二个数据框缺少一些日期。我想这是我尝试绘制它时的问题（数据示例）：

from io import StringIO
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

df_senti = '        date  senti-pol  senti-vader\n0 2019-10-01  -0.060639    -0.174223\n1 2019-10-02  -0.080265     0.090761\n2 2019-10-03  -0.186335    -0.645464\n3 2019-10-04   0.014124    -0.043164\n4 2019-10-05  -0.035157     0.275379'
df_daily = pd.read_csv(StringIO(df_senti),sep='\s+')
print(df_daily)

import yfinance as yf
df_dbk = yf.download("DBK.DE", start="2019-10-02", end="2019-10-10", interval="1d") # day+1 otherwise wrong data
print(df_dbk)

fig, ax1 = plt.subplots()

color = 'tab:red'
ax1.set_xlabel('date')
ax1.set_ylabel('sentiment', color=color)
ax1.plot(df_daily['date'], df_daily['senti-vader'], color=color)
ax1.tick_params(axis='y', labelcolor=color)

ax2 = ax1.twinx()  # instantiate a second axes that shares the same x-axis

color = 'tab:blue'
ax2.set_ylabel('share price', color=color)  # we already handled the x-label with ax1
ax2.plot(df_dbk['Date'], df_dbk['Adj Close'], color=color)
ax2.tick_params(axis='y', labelcolor=color)

fig.tight_layout()  # otherwise the right y-label is slightly clipped
plt.show()

我收到的错误是

KeyError: 'Date'

During handling of the above exception, another exception occurred:

KeyError                                  Traceback (most recent call last)
<ipython-input-11-e4156bac397f> in <module>
     24 color = 'tab:blue'
     25 ax2.set_ylabel('share price', color=color)  # we already handled the x-label with ax1
---> 26 ax2.plot(df_dbk['Date'], df_dbk['Adj Close'], color=color)
     27 ax2.tick_params(axis='y', labelcolor=color)

有没有一种很好的方法可以忽略第二个 DataFrame 中缺失的日期？

Answer 1

'Date' 不是数据框 dbk_df_csv 中的列。它似乎是您的索引的名称。这就是你KeyError 告诉你的。

试试这个

ax2.plot(dbk_df_csv.index, dbk_df_csv['Adj Close'], color=color)

事实上，你可以省略 x 值，它会从数据中推断出来：

ax2.plot(dbk_df_csv['Adj Close'], color=color)

Answer 2

我找到了解决问题的方法。问题在于不同类型的 x 值。

print(type(df_daily.index))
<class 'pandas.core.indexes.numeric.Int64Index'>

print(type(df_daily['date']))
<class 'pandas.core.series.Series'>

print(type(df_dbk.index))
<class 'pandas.core.indexes.datetimes.DatetimeIndex'>

根据@busybear 的回答，我想将索引从df_daily 转换。

df_daily['Date'] = pd.to_datetime(df_daily['date'])
df_daily = df_daily.set_index('Date')

print(df_daily)
                 date  senti-pol  senti-vader
Date                                         
2019-10-01 2019-10-01  -0.060639    -0.174223
2019-10-02 2019-10-02  -0.080265     0.090761
2019-10-03 2019-10-03  -0.186335    -0.645464
2019-10-04 2019-10-04   0.014124    -0.043164
2019-10-05 2019-10-05  -0.035157     0.275379

print(type(df_daily.index))
<class 'pandas.core.indexes.datetimes.DatetimeIndex'>

通过更改索引，我能够生成绘图。

fig, ax1 = plt.subplots(figsize=(20, 10))

color = 'tab:red'
ax1.set_xlabel('date')
ax1.set_ylabel('sentiment', color=color)
ax1.plot(df_daily.index, df_daily['senti-vader'], color=color, marker='.')
ax1.tick_params(axis='y', labelcolor=color)
ax1.tick_params(axis='x', rotation=45)

ax2 = ax1.twinx()  # instantiate a second axes that shares the same x-axis

color = 'tab:blue'
ax2.set_ylabel('share price', color=color)  # we already handled the x-label with ax1
ax2.plot(df_dbk.index, df_dbk['Adj Close'], color=color, marker='.')
ax2.tick_params(axis='y', labelcolor=color)
ax2.tick_params(axis='x', rotation=45)

fig.tight_layout()  # otherwise the right y-label is slightly clipped

plt.show()