Exploring Matplotlib Styles for Visualizations
Take a look at all the available Matplotlib Visualization styles in one-go, for use as a template when needed!
Published on March 03, 2022 by Vimal Octavius PJ
Matplotlib Visualization Aesthetics Time Series Plot
221 min READ
Exploring various styles for visualizations in Matplotlib¶
import matplotlib.pyplot as plt
%matplotlib inline
import warnings
warnings.filterwarnings('ignore')
import numpy as np
Viewing the available styles:
styles = plt.style.available[:]
n = 1
for i in styles:
print(n, i)
n+=1
1 Solarize_Light2 2 _classic_test_patch 3 bmh 4 classic 5 dark_background 6 fast 7 fivethirtyeight 8 ggplot 9 grayscale 10 seaborn 11 seaborn-bright 12 seaborn-colorblind 13 seaborn-dark 14 seaborn-dark-palette 15 seaborn-darkgrid 16 seaborn-deep 17 seaborn-muted 18 seaborn-notebook 19 seaborn-paper 20 seaborn-pastel 21 seaborn-poster 22 seaborn-talk 23 seaborn-ticks 24 seaborn-white 25 seaborn-whitegrid 26 tableau-colorblind10
Plotting the same graph in each style:
series = np.random.randn(50) #Generate random time series data
def trend_graph():
fig = plt.figure(figsize=(13,5))
ax = fig.add_subplot(111)
plt.grid()
ax.plot(series)
ax.set_title('A Time Series Plot', fontsize=18, fontweight='bold')
ax.set_xlabel('Day', fontsize=15)
ax.set_ylabel('Amount', fontsize=15)
ax.axis([0,50, -5, 5])
for s in styles:
with plt.style.context(s):
trend_graph()
I always come back to this code, to select the style I could apply easily to any plot, without much change to the code. All I have to do is:
- Declare a function for my plots
- Check out which plot I like as needed/per context
- Eg., I would select the plot with style number 9 or 'grayscale' if I need a printout. To do so, I just include the following after the function:
- with plt.style.context('grayscale')
- Eg., I would select the plot with style number 9 or 'grayscale' if I need a printout. To do so, I just include the following after the function:
Here goes the code:
series = np.random.randn(50) #Generate random time series data
def trend_graph():
fig = plt.figure(figsize=(13,5))
ax = fig.add_subplot(111)
plt.grid()
ax.plot(series)
ax.set_title('A Time Series Plot', fontsize=18, fontweight='bold')
ax.set_xlabel('Day', fontsize=15)
ax.set_ylabel('Amount', fontsize=15)
ax.axis([0,50, -5, 5])
with plt.style.context('grayscale'):
trend_graph()
If I need a style with a dark background, I have this: (Note that there is no change in the code except the line to display the style "with plt.style.context('dark_background')"
series = np.random.randn(50) #Generate random time series data
def trend_graph():
fig = plt.figure(figsize=(13,5))
ax = fig.add_subplot(111)
plt.grid()
ax.plot(series)
ax.set_title('A Time Series Plot', fontsize=18, fontweight='bold')
ax.set_xlabel('Day', fontsize=15)
ax.set_ylabel('Amount', fontsize=15)
ax.axis([0,50, -5, 5])
with plt.style.context('dark_background'):
trend_graph()
