Better visualization of Pie charts by MatPlotLib

We as humans always like to interpret pictures more than text. This is supported by the biological fact that 90% of facts that come to the brain are in form of Visuals.

When it comes to Data, visualization is a key concept. We grasp more by seeing visuals(graphs, charts, etc.) rather than just seeing data in the form of raw text. Visualization through images makes our brain to process faster and looks pleasant rather than just a big junk of text.

even Irfan likes graphs ;)

When it comes to making graphs, we should make it in such a way that it feels pleasant to mind. A lot of people make data visualization without considering the most important factor “Data-Ink Ratio” — term coined by Edward Rolf Tufte in his writings Information design.

Below GIF taught us the whole concept of the Data-ink ratio.

All-in-one image for understanding data-ink ratio (Source: Darkhorse Analytics)

Pretty Simple, Huh ?!

Now let’s start making our Pie chart — a good looking Pie chart. let’s directly work on an example given in matplotlib documentation:

import matplotlib.pyplot as plt

# Pie chart
labels = ['Frogs', 'Hogs', 'Dogs', 'Logs']
sizes = [15, 30, 45, 10]
# only "explode" the 2nd slice (i.e. 'Hogs')
explode = (0, 0.1, 0, 0)  
fig1, ax1 = plt.subplots()
ax1.pie(sizes, explode=explode, labels=labels, autopct='%1.1f%%',
        shadow=True, startangle=90)
# Equal aspect ratio ensures that pie is drawn as a circle
ax1.axis('equal')  
plt.tight_layout()
plt.show()

The above code gives you a pie chart as below:

Let’s make this chart into a good looking chart:

1. Always use good colors (preferably light shades)
2. Change labels, percent label color
3. Change positions of labels and percent labels
4. Change your pie chart to donut chart(optional)

Let’s add color to graph :

Colors can be added to graph by one of the following ways:

• Hex value (#_ _ _ _ _ _)
• rgb value ( r , g , b ) or ( r , g , b , a ) where r, g, b, a are in range 0–1
• grey shades (0–1 range)
• as in active color cycle ( “c0”, “c1”)[capital C followed by digit]

import matplotlib.pyplot as plt

# Pie chart
labels = ['Frogs', 'Hogs', 'Dogs', 'Logs']
sizes = [15, 30, 45, 10]
# only "explode" the 2nd slice (i.e. 'Hogs')
explode = (0, 0.1, 0, 0)
#add colors
colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99']
fig1, ax1 = plt.subplots()
ax1.pie(sizes, explode=explode, labels=labels, colors=colors, autopct='%1.1f%%',
        shadow=True, startangle=90)
# Equal aspect ratio ensures that pie is drawn as a circle
ax1.axis('equal')
plt.tight_layout()
plt.show()

Changing the color of labels on the chart

We can change the color of labels and percent labels by set_color() property of matplotlib.text.Text object which are return type of function plot.pie().

Modify the above code as below, you can find more info about the pie function’s return datatype here. Also, let’s turn off shadow effect and explode effect.

import matplotlib.pyplot as plt
# Pie chart
labels = ['Frogs', 'Hogs', 'Dogs', 'Logs']
sizes = [15, 30, 45, 10]
#colors
colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99']
 
fig1, ax1 = plt.subplots()
patches, texts, autotexts = ax1.pie(sizes, colors = colors, labels=labels, autopct='%1.1f%%', startangle=90)
for text in texts:
    text.set_color('grey')
for autotext in autotexts:
    autotext.set_color('grey')
# Equal aspect ratio ensures that pie is drawn as a circle
ax1.axis('equal')  
plt.tight_layout()
plt.show()

Now our pie chart looks something like this:

color changes of labels

Changing the pie chart to donut chart to look more cool and AWESOME!!

Method of making donut graph from pie chart in MatPlotLib documentation is complicated rather we can just change the pie chart to donut chart by drawing a circle with white color at origin(Source: here).

Modify code as below to draw a circle centered at (0,0)

import matplotlib.pyplot as plt
# Pie chart
labels = ['Frogs', 'Hogs', 'Dogs', 'Logs']
sizes = [15, 30, 45, 10]
#colors
colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99']
 
fig1, ax1 = plt.subplots()
ax1.pie(sizes, colors = colors, labels=labels, autopct='%1.1f%%', startangle=90)
#draw circle
centre_circle = plt.Circle((0,0),0.70,fc='white')
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
# Equal aspect ratio ensures that pie is drawn as a circle
ax1.axis('equal')  
plt.tight_layout()
plt.show()

donut chart from pie chart with white circle at center

Changing positions of labels

We can change the position of labels (both outer and percent labels) by modifying labeldistance(defaul:1) and pctdistance(default:0.6)

Let’s modify code by adding pctdistance=0.85 and use explosion property to make it looks better.

import matplotlib.pyplot as plt
# Pie chart
labels = ['Frogs', 'Hogs', 'Dogs', 'Logs']
sizes = [15, 30, 45, 10]
#colors
colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99']
#explsion
explode = (0.05,0.05,0.05,0.05)
 
plot.pie(sizes, colors = colors, labels=labels, autopct='%1.1f%%', startangle=90, pctdistance=0.85, explode = explode)
#draw circle
centre_circle = plt.Circle((0,0),0.70,fc='white')
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
# Equal aspect ratio ensures that pie is drawn as a circle
ax1.axis('equal')  
plt.tight_layout()
plt.show()

Donut chart with explode property

Let’s work on another example and see how we can implement multiple-pie charts to make data more visualizable.

Pie charts for the above data will be as follows:

Instead of visualizing different pie charts of gender for different languages, we can implement all the pie charts in one chart as given below:

import matplotlib.pyplot as plt
 
# Data to plot
labels = ['Python', 'C++', 'Ruby', 'Java']
sizes = [504, 337, 415, 280]
labels_gender = ['Man','Woman','Man','Woman','Man','Woman','Man','Woman']
sizes_gender = [315,189,125,212,270,145,190,90]
colors = ['#ff6666', '#ffcc99', '#99ff99', '#66b3ff']
colors_gender = ['#c2c2f0','#ffb3e6', '#c2c2f0','#ffb3e6', '#c2c2f0','#ffb3e6', '#c2c2f0','#ffb3e6']
 
# Plot
plt.pie(sizes, labels=labels, colors=colors, startangle=90,frame=True)
plt.pie(sizes_gender,colors=colors_gender,radius=0.75,startangle=90)
centre_circle = plt.Circle((0,0),0.5,color='black', fc='white',linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
 
plt.axis('equal')
plt.tight_layout()
plt.show()

Or we can just use the explosion property of slices of a pie chart to make it more cool..

import matplotlib.pyplot as plt
 
# Data to plot
labels = ['Python', 'C++', 'Ruby', 'Java']
sizes = [504, 337, 415, 280]
labels_gender = ['Man','Woman','Man','Woman','Man','Woman','Man','Woman']
sizes_gender = [315,189,125,212,270,145,190,90]
colors = ['#ff6666', '#ffcc99', '#99ff99', '#66b3ff']
colors_gender = ['#c2c2f0','#ffb3e6', '#c2c2f0','#ffb3e6', '#c2c2f0','#ffb3e6', '#c2c2f0','#ffb3e6']
explode = (0.2,0.2,0.2,0.2) 
explode_gender = (0.1,0.1,0.1,0.1,0.1,0.1,0.1,0.1)
#Plot
plt.pie(sizes, labels=labels, colors=colors, startangle=90,frame=True, explode=explode,radius=3)
plt.pie(sizes_gender,colors=colors_gender,startangle=90, explode=explode_gender,radius=2 )
#Draw circle
centre_circle = plt.Circle((0,0),1.5,color='black', fc='white',linewidth=0)
fig = plt.gcf()
fig.gca().add_artist(centre_circle)
 
plt.axis('equal')
plt.tight_layout()
plt.show()

Looking cool, right ?!!

Thanks for reading this article, recommend and share if you like it. You can even buy me a cup of coffee(https://buymeacoff.ee/kvnamipara).