Python Matplotlib

Table of Contents:

Set figure size and dpi
Creating sublots
Graphs labeling
Subplots labeling
Multiple plots
Multiple scatter
Different scatter color
Different scatter size
Multiple subplots
Lines
Pcolor and Pcolormesh

What would be the crucial knowledge to successfully create matplotlib based graphs?

Set figure size and dpi

First you can control the figure size and dpi quality of the plot.

import matplotlib.pyplot as plt
plt.figure(figsize=(3, 4), dpi=50)

Bigger figsize and dpi,bigger the plot will be.

Creating sublots

matplotlib new api can now create subplots. You can use add_subplot method for this purpose:

import matplotlib.pyplot as plt
fig = plt.figure()
fig.add_subplot(2,2,1)
plt.show()

Associated method gca may be handy:

ax=plt.gca()

gca means “get current axes”, and it may help you find the “current” or last active axes.

If there is no axes yet, an axes will be created. If you create two subplots, the subplot that is created last is the current one.

Instead of getting the ax handle like above, it is smarter to create handles at once:

import matplotlib.pyplot as plt
fig = plt.figure()
ax1 = fig.add_subplot(2,2,1)
plt.show()

This way you don’t need to use gca() method. You have another alternative to create subplots with the subplots method. Following two examples will provide equivalent results:

Example 1: Using subplots

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)
fig, (ax) = plt.subplots(ncols=2, nrows=2)
plt.rcParams["figure.figsize"] = (9,6)
plt.rcParams["figure.dpi"] = 200
# or plt.figure(figsize=(9, 6), dpi=200)

ax1=ax[0][0]
ax2=ax[0][1]
ax3=ax[1][0]
ax4=ax[1][1]
ax1.scatter(t,a,c='c')
ax2.scatter(t,a,c='g')
ax3.scatter(t,a,c='r')
ax4.scatter(t,a,c='m')
plt.show()

Example 2: Using add_subplot

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)

fig = plt.figure(figsize=(9, 6), dpi=200)
ax1=fig.add_subplot(2,2,1)   
ax1.scatter(t,a,c='c')
ax2=fig.add_subplot(2,2,2)  
ax2.scatter(t,a,c='g')
ax3=fig.add_subplot(2,2,3)  
ax3.scatter(t,a,c='r')
ax4=fig.add_subplot(2,2,4)  
ax4.scatter(t,a,c='m')

plt.show()

Graphs labeling

If you don’t use subplots you can use old api:

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)
fig = plt.figure(figsize=(9, 6), dpi=150)
plt.title("title")
plt.xlabel("x axis label")
plt.ylabel("y axis label")
plt.scatter(t,a)
plt.show()

plots label

Subplots labeling

Subplots uses new api for labels like this:

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)
fig = plt.figure(figsize=(9, 6), dpi=150)

ax1=fig.add_subplot(1,3,1)   
ax1.title.set_text('First Plot')
ax1.set_xlabel('first xlabel')
ax1.set_ylabel('first ylabel')
ax1.scatter(t,a,c='c')

ax2=fig.add_subplot(3,3,2)
ax2.title.set_text('Second Plot')
ax2.set_xlabel('second xlabel')
ax2.set_ylabel('second ylabel')
ax2.scatter(t,a,c='g')

plt.tight_layout()
plt.show()

subplots label

With the new subplots idea and labels idea for subplots you can have this boilerplate:

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)
fig = plt.figure(figsize=(9, 6), dpi=150)
ax1=fig.add_subplot(1,1,1)  
ax1.title.set_text('title')
ax1.set_xlabel('x axis label')
ax1.set_ylabel('y axis label')
ax1.scatter(t,a)
plt.show()

plots label

Multiple plots

Adding multiple plots to the same plot is possible:

import numpy as np
from matplotlib import pyplot as plt
t = np.linspace(0, 2*math.pi, 800)
a = np.sin(t)
b = np.cos(t)
plt.figure(figsize=(9, 6), dpi=75)
plt.plot(t, a, 'r') # red
plt.plot(t, b, 'g') # green
plt.show()

multiple plots

However, I would always use this way:

import numpy as np
from matplotlib import pyplot as plt
t = np.linspace(0, 2*math.pi, 800)
a = np.sin(t)
b = np.cos(t)
fig = plt.figure(figsize=(9, 6), dpi=150)
ax1=fig.add_subplot(1,1,1)  
ax1.title.set_text('title')
ax1.set_xlabel('x label')
ax1.set_ylabel('y label')
ax1.plot(t, a, 'r') # red
ax1.plot(t, b, 'g') # green
plt.show()

We used in here the numpy sin and cos functions, these are different from the import math sin and cos functions that work on scalars only.

Multiple scatter

Here we combine two scatter plots like before we combined several plots.

import math as m
import numpy as np

t = np.linspace(0, 2*math.pi, 40)
a = np.sin(t)
b = np.cos(t)
c = a + b

plt.scatter(t, a, c='r')
plt.scatter(t, b, c='b')
plt.show()

multiple scatter

Different scatter color

from matplotlib import pyplot as plt
plt.figure(figsize=(5, 5), dpi=150)
plt.title('stars')
co = ['g']+['r', 'g', 'b']*33 
plt.scatter(train_data[:,0], train_data[:,1], s=55, c=co, marker="*")
plt.scatter(train_data[:,1], train_data[:,0], s=1, c='c')
plt.show()

multiple scatter

Different scatter size

import numpy as np
import matplotlib.pyplot as plt
np.random.seed(1001) # to reproduce
N = 150
x = np.random.rand(N)
y = np.random.rand(N)
colors = np.random.rand(N)
size = (20 * np.random.rand(N))**2  
plt.figure(figsize=(9, 6), dpi=75)
plt.scatter(x, y, s=size, c=colors, alpha=0.5)
plt.show()

scatter size

As we insist on using the subplots the code should better be like this:

import numpy as np
import matplotlib.pyplot as plt
np.random.seed(1001) # to reproduce
N = 150
x = np.random.rand(N)
y = np.random.rand(N)
colors = np.random.rand(N)
size = (20 * np.random.rand(N))**2  
fig = plt.figure(figsize=(9, 6), dpi=150)
ax1=fig.add_subplot(1,1,1)  
ax1.title.set_text('title')
ax1.set_xlabel('x label')
ax1.set_ylabel('y label')
ax1.scatter(x, y, s=size, c=colors, alpha=0.5)
plt.show()

Multiple subplots

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)
fig = plt.figure(figsize=(9, 6), dpi=100)
ax1=fig.add_subplot(2,2,1)   

ax1.scatter(t,a,c='c')
ax2=fig.add_subplot(222)  
ax2.scatter(t,a,c='g')
ax3=fig.add_subplot(223)  
ax3.scatter(t,a,c='r')
ax4=fig.add_subplot(224)  
ax4.scatter(t,a,c='m')
plt.show()

4 subplots

If you would remove the third subplot and make the first one double in size this would look like:

import matplotlib.pyplot as plt
t = np.linspace(0, 2*math.pi, 60)
a = np.sin(t)
fig = plt.figure(figsize=(9, 6), dpi=100)
ax1=fig.add_subplot(1,2,1)   
ax1.scatter(t,a,c='c')
ax2=fig.add_subplot(2,2,2)  
ax2.scatter(t,a,c='g')
ax4=fig.add_subplot(2,2,4)  
ax4.scatter(t,a,c='m')
plt.show()

3 subplots

Lines

Example:

import matplotlib.pyplot as plt
fig, ax = plt.subplots(figsize = (7,5), dpi=150 )
ax.vlines(5,1,2,colors='r')
ax.vlines(3,1,2,colors='r')
ax.vlines(4,0,3,colors='r')
ax.hlines(5,1,2,colors='g')
ax.hlines(2,0,3,colors='g')
ax.hlines(3,1,2,colors='g')
x1, y1 = [-1, 5], [1, 4]
ax.plot(x1, y1, marker = 'o')
plt.show()

Output: lines

Pcolor and Pcolormesh

import matplotlib.pyplot as plt
import numpy as np

Z = np.random.rand(6, 10)
fig, (ax0) = plt.subplots(1, 1)
c = ax0.pcolor(Z)
ax0.set_title('pcolor')
plt.show()

We can just have the random values and using pcolor or pcolormesh we can draw the mesh.

pcolor

However, the more realistic is to create the grid.

import matplotlib.pyplot as plt
import numpy as np

dx, dy = 0.05, 0.05
y, x = np.mgrid[slice(1, 5 + dy, dy),
                slice(1, 5 + dx, dx)]
z = np.sin(x)**10 + np.cos(10 + y*x) * np.cos(x)
fig, (ax0) = plt.subplots(1, 1)
c = ax0.pcolormesh(x,y,z)
ax0.set_title('pcolormesh')
plt.show()

pcolormesh

…

tags: random forests & category: python