Python 从两列中绘制熊猫数据框

plot pandas dataframe two columns from

提问by Alessandro Sorvillo

i have a pandas dataframe which has dates as indexes and some columns: I would like to plot a line chart with 2 lines (let's say 'ISP.MI' and 'Ctrv'); on the x axis I need the 'Date'

我有一个 Pandas 数据框，它有日期作为索引和一些列：我想用 2 行绘制折线图（比如“ISP.MI”和“Ctrv”）；在 x 轴上我需要“日期”

Ticker       ISP.MI  Daily returns        Ctrv  Inv_Am  Giac_Media
Date                                                                 
2016-01-01  2.90117            NaN  100.000000     100       100.0   
2016-01-04  2.80159      -0.034927  196.507301     200       150.0   
2016-01-05  2.85608       0.019263  300.292610     300       200.0   
2016-01-06  2.77904      -0.027345  392.081255     400       250.0   
2016-01-07  2.73206      -0.017050  485.396411     500       300.0   
2016-01-08  2.72267      -0.003443  583.725246     600       350.0   

采纳答案by epattaro

if you dont care about axis scale:

如果您不关心轴比例：

plt.figure()

x = df['Date']
y1 = df['ISP.MI']
y2 = df['Ctrv']

plt.plot(x,y1)
plt.plot(x,y2)

if you do care about it:

如果你真的关心它：

fig, ax1 = plt.subplots()

x = df['Date']
y1 = df['ISP.MI']
y2 = df['Ctrv']

ax2 = ax1.twinx()

ax1.plot(x, y1, 'g-')
ax2.plot(x, y2, 'b-')

回答by jezrael

I think the simpliest is select columns by subset and then DataFrame.plot:

我认为最简单的是按子集选择列，然后DataFrame.plot：

df[['ISP.MI','Ctrv']].plot()

回答by alexbhandari

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

d = {'x' : [1,2,3,4,5,6,7,8,9,10],
     'y_one' : np.random.rand(10),
     'y_two' : np.random.rand(10)}

df = pd.DataFrame(d)

df.plot('x',y=['y_one','y_two'])
plt.show()

回答by Pezze

So, here is the code that from scratch creates a dataframe that looks like yours and generates the plot you asked for:

因此，这是从头开始创建一个看起来像您的数据框并生成您要求的图的代码：

import pandas as pd
import datetime
import numpy as np
from matplotlib import pyplot as plt

# The following two lines are not mandatory for the code to work
import matplotlib.style as style
style.use('dark_background')

def create_datetime_range(numdays=10):
    """Creates the timestamp range"""
    base = datetime.datetime.today()
    datelist = pd.date_range(base, periods=numdays).to_pydatetime()
    return datelist
def convert_to_date(datetime_list):
    """Converts a timestamp array into a date array"""
    return [x.date() for x in datetime_list]



a = pd.DataFrame(
    {
        'ISP.MI': np.random.normal(2,1,10),
        'Ctrv' : np.random.normal(200,150,10)
    }, 
    index=convert_to_date(create_date_range())
)
a.plot()

However, I believe that your dataframe is different in two ways:

但是，我相信您的数据框在两个方面有所不同：

1. It seems that there are two levels in the index (the Date title seems on a second row to the Ticker title). I imagine this could be because you used something like .groupby() or .unstack() or other aggregation/pivoting method. I suggest you to look in the reset_index() method.

1. 索引中似乎有两个级别（日期标题似乎在代码标题的第二行）。我想这可能是因为您使用了 .groupby() 或 .unstack() 或其他聚合/旋转方法。我建议您查看 reset_index() 方法。

2.Your dataframe has more columns that you need. As suggested by @jezrael, you should first select only these. You can do it with something like:

2.您的数据框有更多您需要的列。正如@jezrael 所建议的，您应该首先只选择这些。你可以这样做：

df[['ISP.MI','Ctrv']]

and then using the .plot() method on the smaller dataframe and let pandas handle the rest.

然后在较小的数据帧上使用 .plot() 方法，让 Pandas 处理剩下的事情。

回答by shantanu pathak

Now in latest pandas you can directly use df.plot.scatter function

现在在最新的 Pandas 中你可以直接使用 df.plot.scatter 函数

df = pd.DataFrame([[5.1, 3.5, 0], [4.9, 3.0, 0], [7.0, 3.2, 1],
               [6.4, 3.2, 1], [5.9, 3.0, 2]],
              columns=['length', 'width', 'species'])
ax1 = df.plot.scatter(x='length',
                  y='width',
                  c='DarkBlue')

https://pandas.pydata.org/pandas-docs/version/0.23/generated/pandas.DataFrame.plot.scatter.html

https://pandas.pydata.org/pandas-docs/version/0.23/generated/pandas.DataFrame.plot.scatter.html