Plot a bar plot from a Pandas DataFrame

Question

Assuming I have a DataFrame that looks like this:

Hour  V1  V2  A1  A2
   0  15  13  25  37
   1  26  52  21  45
   2  18  45  45  25
   3  65  38  98  14

I'm trying to create a bar plot to compare columns V1 and V2 by the Hour. When I do:

import matplotlib.pyplot as plt
ax = df.plot(kind='bar', title ="V comp",figsize=(15,10),legend=True, fontsize=12)
ax.set_xlabel("Hour",fontsize=12)
ax.set_ylabel("V",fontsize=12)

I get a plot and a legend with all the columns' values and names. How can I modify my code so the plot and legend only displays the columns V1 and V2?

Would it not deliver what you want by just sub-selecting the columns of interest first: ax = df[['V1','V2']].plot(kind='bar', title ="V comp",figsize=(15,10),legend=True, fontsize=12)? — EdChum
– EdChum, Commented Apr 7, 2015 at 18:41
@EdChum I was trying along those line but i kept getting error messages. I'm using pandas.pydata.org/pandas-docs/version/0.15.2/visualization.html as reference and could not find an example of what i needed. But you are correct. If you answer it i can accept it. If you can add the explanation of the [ [ ] ] syntax since i was doing ax = df['V1','V2'] — Alfonso
– Alfonso, Commented Apr 7, 2015 at 18:45

johnbaltis · Accepted Answer · 2016-10-28 15:38:48Z

84

To plot just a selection of your columns you can select the columns of interest by passing a list to the subscript operator:

ax = df[['V1','V2']].plot(kind='bar', title ="V comp", figsize=(15, 10), legend=True, fontsize=12)

What you tried was df['V1','V2'] this will raise a KeyError as correctly no column exists with that label, although it looks funny at first you have to consider that your are passing a list hence the double square brackets [[]].

import matplotlib.pyplot as plt
ax = df[['V1','V2']].plot(kind='bar', title ="V comp", figsize=(15, 10), legend=True, fontsize=12)
ax.set_xlabel("Hour", fontsize=12)
ax.set_ylabel("V", fontsize=12)
plt.show()

enter image description here

edited Oct 28, 2016 at 15:38

johnbaltis

1,5874 gold badges18 silver badges29 bronze badges

answered Apr 7, 2015 at 18:56

EdChum

397k204 gold badges836 silver badges583 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

FaCoffee Over a year ago

Ok, how would you add the relevant error bars to each series?

EdChum Over a year ago

@CF84 you should post a question rather than ask a question via comments as this doesn't help the SO community

San Askaruly · Accepted Answer · 2022-08-09 11:16:58Z

1

A minimal seaborn apporach to plot from pandas dataframe:

import pandas as pd
import seaborn as sns

random_dict = {
    "classes": ["A", "B", "C"],
    "quantities": [100, 300, 200]
}

df = pd.DataFrame.from_dict(random_dict)
sns.barplot(x="classes", y="quantities", data=df)

answered Aug 9, 2022 at 11:16

San Askaruly

3411 silver badge11 bronze badges

Comments

cottontail · Accepted Answer · 2023-09-26 05:36:48Z

Column labels may be passed as axis labels (y-axis admits multiple column labels). Also, since pandas 1.1, axis labels may be passed into the plot() call as well.

df.plot(x='Hour', y=['V1', 'V2'], kind='bar', title="V comp", figsize=(12,6), ylabel='V', rot=0);

Another example where bar colors are set and bars are labeled by its height (must have matplotlib>=3.4).

ax = df.plot(x='Hour',                # values on x-axis
             y=['V1', 'V2'],          # values on y-axis
             kind='bar',              # specify that it is a bar-plot
             title="V comp",          # set title
             figsize=(12,6),          # set figure size
             ylabel='V',              # set y-axis label
             rot=0,                   # do not rotate x-ticklabels
             color=['red', 'blue'])   # set bar colors
for heights in ax.containers:
    ax.bar_label(heights)             # label each bar by its height

Collectives™ on Stack Overflow

Plot a bar plot from a Pandas DataFrame

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related