matplotlib: assigning different hatch to bars

Question

I have a dataframe where for each index, I have to plot two bars (for two series). The following code gives the output as:

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

df = pd.DataFrame(np.random.randint(0,20,size=(5, 2)), columns=list('AB'))
fig, ax = plt.subplots()
ax = df.sort_values('B', ascending=True).plot.barh(rot=0,ax=ax,hatch="/")
plt.show()

I would like to assign individual hatching for each bar. So that if A has ‘/’ hatching, B should have ‘|’. What modifications do I need to make in the code?

Asked By: SaadH

||

Source

Answer 1

You can plot the two bars separately:

import numpy as np
import pandas as pd

from matplotlib import pyplot as plt

df = pd.DataFrame(np.random.randint(0, 20, size=(5, 2)), columns=list('AB'))
fig, ax = plt.subplots()

ax.barh(np.arange(0, len(df)), df['A'], height=0.3, hatch='/')
ax.barh(np.arange(0.3, len(df) + 0.3), df['B'], height=0.3, hatch='|')

Answered By: gmds

Answer 2

The matplotlib example features a solution. But I don’t really like it, since it is aiming to set a different hatch to each bar.

But imho in most cases it is more relevant to set a specific hatch to each "category" of bar. You can either do it by plotting the bars separately with a hatch, or set the hatches after plotting. Setting the hatches after plotting is imho much more flexible, thus this is my approach:

df = pd.DataFrame(np.random.randint(0,20,size=(5, 2)), columns=list('AB'))
fig, ax = plt.subplots()
ax = df.sort_values('B', ascending=True).plot.barh(rot=0,ax=ax)
# get all bars in the plot
bars = ax.patches
patterns = ['/', '|']  # set hatch patterns in the correct order
hatches = []  # list for hatches in the order of the bars
for h in patterns:  # loop over patterns to create bar-ordered hatches
    for i in range(int(len(bars) / len(patterns))):
        hatches.append(h)
for bar, hatch in zip(bars, hatches):  # loop over bars and hatches to set hatches in correct order
    bar.set_hatch(hatch)
# generate legend. this is important to set explicitly, otherwise no hatches will be shown!
ax.legend()
plt.show()

The advantages of this solution over plotting the bars separately are:

you can have an arbitrary number of bars
works with stacked and/or unstacked bars in all possible combinations
works with the pandas plotting interface

With the main drawback being the additional LOC especially for plotting only a few bars. But packing it into a function/module and re-using it solves this problem. 🙂

Answered By: JE_Muc

Answer 3

here is a test ca can help you

import numpy as np
import matplotlib.pyplot as plt

df = pd.DataFrame(np.random.randint(0,20,size=(5, 2)), columns=list('AB'))

plt.hist(df['A'], color = 'blue',
            edgecolor = 'red', hatch = '/' , label = 'df.A',orientation = 'horizontal',
            histtype = 'bar')
plt.hist(df['B'],color = 'YELLOW',
            edgecolor = 'GREEN', hatch = 'O' , label = 'df.B',orientation = 'horizontal',
            histtype = 'bar')
plt.legend()
plt.show()

Answered By: Oussama Bamaàrouf

matplotlib: assigning different hatch to bars

Question:

Answers: