Python pandas，绘制多行的选项

Question

I want to plot multiple lines from a pandas dataframe and setting different options for each line. I would like to do something like

testdataframe=pd.DataFrame(np.arange(12).reshape(4,3))
testdataframe.plot(style=['s-','o-','^-'],color=['b','r','y'],linewidth=[2,1,1])

This will raise some error messages:

• linewidth is not callable with a list

• In style I can't use 's' and 'o' or any other alphabetical symbol, when defining colors in a list

Also there is some more stuff which seems weird to me

• when I add another plot command to the above code testdataframe[0].plot() it will plot this line in the same plot, if I add the command testdataframe[[0,1]].plot() it will create a new plot

• If i would call testdataframe[0].plot(style=['s-','o-','^-'],color=['b','r','y']) it is fine with a list in style, but not with a list in color

Hope somebody can help, thanks.

Answer 1

You're so close!

You can specify the colors in the styles list:

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

testdataframe = pd.DataFrame(np.arange(12).reshape(4,3), columns=['A', 'B', 'C'])
styles = ['bs-','ro-','y^-']
linewidths = [2, 1, 4]
fig, ax = plt.subplots()
for col, style, lw in zip(testdataframe.columns, styles, linewidths):
    testdataframe[col].plot(style=style, lw=lw, ax=ax)

Also note that the plot method can take a matplotlib.axes object, so you can make multiple calls like this (if you want to):

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

testdataframe1 = pd.DataFrame(np.arange(12).reshape(4,3), columns=['A', 'B', 'C'])
testdataframe2 = pd.DataFrame(np.random.normal(size=(4,3)), columns=['D', 'E', 'F'])
styles1 = ['bs-','ro-','y^-']
styles2 = ['rs-','go-','b^-']
fig, ax = plt.subplots()
testdataframe1.plot(style=styles1, ax=ax)
testdataframe2.plot(style=styles2, ax=ax)

Not really practical in this case, but the concept might come in handy later.

Answer 2

Considering the dataframe testdataframe

testdataframe = pd.DataFrame(np.arange(12).reshape(4,3))

print(testdataframe)

   0   1   2
0  0   1   2
1  3   4   5
2  6   7   8
3  9  10  11

You can combine styles into a single list of strings as in styles defined below. I'll also define the linewidths in lws

styles=['bs-', 'ro-', 'y^-']
lws = [2, 1, 1]

We can use the plot method on the testdataframe passing the list styles to the style parameter. Note that we could have also passed a dictionary (and probably other things as well).

However, line widths are not as easily handled. I first capture the AxesSubplot object and iterate over the lines attribute setting the line width.

ax = testdataframe.plot(style=styles)
for i, l in enumerate(ax.lines):
    plt.setp(l, linewidth=lws[i])

Answer 3

So I think the answer lies in passing the color and style in the same argument. The following example works with pandas 0.19.2:

testdataframe=pd.DataFrame(np.arange(12).reshape(4,3))
testdataframe.plot(style=['r*-','bo-','y^-'], linewidth=2.0)

Unfortunately, it seems that passing multiple line widths as an input to matplotlib is not possible.