Add x and y labels to a pandas plot

Question

Suppose I have the following code that plots something very simple using pandas:

import pandas as pd
values = [[1, 2], [2, 5]]
df2 = pd.DataFrame(values, columns=['Type A', 'Type B'], 
                   index=['Index 1', 'Index 2'])
df2.plot(lw=2, colormap='jet', marker='.', markersize=10, 
         title='Video streaming dropout by category')

Output

How do I easily set x and y-labels while preserving my ability to use specific colormaps? I noticed that the plot() wrapper for pandas DataFrames doesn't take any parameters specific for that.

This is now the correct [answer](https://stackoverflow.com/a/65875329/7758804) — Trenton McKinney, Aug 14 '21 at 18:33

score 442 · Answer 1 · edited Jan 20 '20 at 20:59

442

The df.plot() function returns a matplotlib.axes.AxesSubplot object. You can set the labels on that object.

ax = df2.plot(lw=2, colormap='jet', marker='.', markersize=10, title='Video streaming dropout by category')
ax.set_xlabel("x label")
ax.set_ylabel("y label")

enter image description here

Or, more succinctly: ax.set(xlabel="x label", ylabel="y label").

Alternatively, the index x-axis label is automatically set to the Index name, if it has one. so df2.index.name = 'x label' would work too.

edited Jan 20 '20 at 20:59

Jaroslav Bezděk

6,967
6
29
46

answered Jan 31 '14 at 18:35

TomAugspurger

28,234
8
86
69

set_xlabel or set_ylabel are not working for pandas 0.25.1. Howerver, ax.set(xlabel="x label", ylabel="y label") does. – erickfis Sep 01 '21 at 19:45

score 64 · Answer 2 · edited Apr 22 '18 at 00:09

64

You can use do it like this:

import matplotlib.pyplot as plt 
import pandas as pd

plt.figure()
values = [[1, 2], [2, 5]]
df2 = pd.DataFrame(values, columns=['Type A', 'Type B'], 
                   index=['Index 1', 'Index 2'])
df2.plot(lw=2, colormap='jet', marker='.', markersize=10,
         title='Video streaming dropout by category')
plt.xlabel('xlabel')
plt.ylabel('ylabel')
plt.show()

Obviously you have to replace the strings 'xlabel' and 'ylabel' with what you want them to be.

edited Apr 22 '18 at 00:09

Max Ghenis

14,783
16
84
132

answered Jan 31 '14 at 18:52

jesukumar

1,139
9
19

1

Also note, you have to call `plt.xlabel()` etc. after `df.plot()`, not before, because otherwise you get two plots - the calls will modify a "previous" plot. Same thing goes for `plt.title()`. – Tomasz Gandor Apr 02 '20 at 17:00

score 33 · Answer 3 · edited Apr 22 '18 at 00:08

33

If you label the columns and index of your DataFrame, pandas will automatically supply appropriate labels:

import pandas as pd
values = [[1, 2], [2, 5]]
df = pd.DataFrame(values, columns=['Type A', 'Type B'], 
                  index=['Index 1', 'Index 2'])
df.columns.name = 'Type'
df.index.name = 'Index'
df.plot(lw=2, colormap='jet', marker='.', markersize=10, 
        title='Video streaming dropout by category')

enter image description here

In this case, you'll still need to supply y-labels manually (e.g., via plt.ylabel as shown in the other answers).

edited Apr 22 '18 at 00:08

Max Ghenis

14,783
16
84
132

answered Dec 17 '14 at 08:06

shoyer

9,165
1
37
55

currently, that 'automatic supply from DataFrame' doesn't work. I just tried it (pandas version 0.16.0, matplotlib 1.4.3) and the plot generates correctly, but with no labels on the axes. – szeitlin Apr 29 '15 at 18:22
1

@szeitlin could you please file a bug report on the pandas github page? https://github.com/pydata/pandas/issues – shoyer Apr 29 '15 at 21:28
you know what, today at least the xlabel is working. maybe there was something strange about the dataframe I was using yesterday (?). if I can reproduce it, I will file it! – szeitlin Apr 30 '15 at 21:27

Serenity · Answer 4 · 2017-05-01T03:56:42.807

30

It is possible to set both labels together with axis.set function. Look for the example:

import pandas as pd
import matplotlib.pyplot as plt
values = [[1,2], [2,5]]
df2 = pd.DataFrame(values, columns=['Type A', 'Type B'], index=['Index 1','Index 2'])
ax = df2.plot(lw=2,colormap='jet',marker='.',markersize=10,title='Video streaming dropout by category')
# set labels for both axes
ax.set(xlabel='x axis', ylabel='y axis')
plt.show()

edited May 01 '17 at 03:56

answered Apr 25 '17 at 12:09

Serenity

35,289
20
120
115

3

I like the `.set(xlabel='x axis', ylabel='y axis')` solution because it lets me put it all in one line, unlike the set_xlabel and set_ylabel plot methods. I wonder why they all (including the set method, by the way) don't return the plot object or at least something inherited from it. – fault-tolerant Oct 28 '17 at 09:02

Mykola Zotko · Accepted Answer · 2021-06-24T17:20:14.330

27

In Pandas version 1.10 you can use parameters xlabel and ylabel in the method plot:

df.plot(xlabel='X Label', ylabel='Y Label', title='Plot Title')

edited Jun 24 '21 at 17:20

answered Jan 24 '21 at 20:03

Mykola Zotko

15,583
3
71
73

score 16 · Answer 6 · answered Apr 05 '17 at 19:22

16

For cases where you use pandas.DataFrame.hist:

plt = df.Column_A.hist(bins=10)

Note that you get an ARRAY of plots, rather than a plot. Thus to set the x label you will need to do something like this

plt[0][0].set_xlabel("column A")

answered Apr 05 '17 at 19:22

Selah

7,728
9
48
60

score 13 · Answer 7 · edited Apr 06 '20 at 08:35

13

what about ...

import pandas as pd
import matplotlib.pyplot as plt

values = [[1,2], [2,5]]

df2 = pd.DataFrame(values, columns=['Type A', 'Type B'], index=['Index 1','Index 2'])

(df2.plot(lw=2,
          colormap='jet',
          marker='.',
          markersize=10,
          title='Video streaming dropout by category')
    .set(xlabel='x axis',
         ylabel='y axis'))

plt.show()

edited Apr 06 '20 at 08:35

Rémy Hosseinkhan Boucher

145
9

answered Sep 18 '19 at 08:20

Dror Hilman

6,837
9
39
56

score 4 · Answer 8 · answered Jun 23 '19 at 10:33

pandas uses matplotlib for basic dataframe plots. So, if you are using pandas for basic plot you can use matplotlib for plot customization. However, I propose an alternative method here using seaborn which allows more customization of the plot while not going into the basic level of matplotlib.

Working Code:

import pandas as pd
import seaborn as sns
values = [[1, 2], [2, 5]]
df2 = pd.DataFrame(values, columns=['Type A', 'Type B'], 
                   index=['Index 1', 'Index 2'])
ax= sns.lineplot(data=df2, markers= True)
ax.set(xlabel='xlabel', ylabel='ylabel', title='Video streaming dropout by category')

This specific use case doesn't seem like a reason to use seaborn. As shown in [the top-voted answer](https://stackoverflow.com/a/21487560/1711796), you can call `set` directly on the value returned from `DataFrame.plot` (which is very similar to the code you've shown here, except without the added dependency). — Bernhard Barker, Dec 16 '20 at 09:53

Add x and y labels to a pandas plot

8 Answers8

Linked

Related