I wrote the basic operation of matplotlib with Jupyter Lab

This article is an article that I actually coded the basic operation of matplotlib described in Kame (@usdatascientist)'s blog (https://datawokagaku.com/python_for_ds_summary/) using Jupyter Lab.

Summary of basic operations of matplotlib

20th

import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

x = np.linspace(-3, 3, 10)
y = np.exp(x)
print(x)
print(y)

[-3.         -2.33333333 -1.66666667 -1.         -0.33333333  0.33333333
  1.          1.66666667  2.33333333  3.        ]
[ 0.04978707  0.09697197  0.1888756   0.36787944  0.71653131  1.39561243
  2.71828183  5.29449005 10.3122585  20.08553692]

plt.plot(x,y)

[<matplotlib.lines.Line2D at 0x7f0518676c50>]

png

Attach attached information

When drawing with matplotlib, you can add (or delete) various attached information. The ones I often use are as follows.

Label the x-axis with plt.xlabel ()

Label the y-axis with plt.ylabel ()

Give the figure a title with plt.title ()

Label the plot with plt.plot (label ='label') and add a legend with plt.legend ()

Add arbitrary ticks to the x-axis with plt.xticks ()

Add arbitrary ticks to the y-axis with plt.yticks ()

Erase the axis with plt.axis (‘off’)

plt.plot(x,y)
# plt.xlabel()Label the x-axis with
plt.xlabel('Efforts')
# plt.ylabel()Label the y-axis with
plt.ylabel('Earning')
# plt.title()Give a title to the figure with
plt.title('This is how your efforts earns')

Text(0.5, 1.0, 'This is how your efforts earns')

png

# plt.plot(label='label')でplotにlabelをつけ, plt.legend()To give a legend
plt.plot(x, y, label='Earning with effort')
plt.legend()
# plt.xticks()Add arbitrary ticks to the x-axis with
plt.xticks(np.arange(-3, 4, 0.5))
# plt.yticks()Attach arbitrary ticks to the y-axis with
plt.yticks([0, 5, 10, 20])
plt.show()

png

x = np.linspace(-3, 3, 10)
y1 = np.exp(x)
y2 = np.exp(x)*2
plt.plot(x, y1, label='first')
plt.plot(x, y2, label='second')
plt.legend()

<matplotlib.legend.Legend at 0x7f05182abc50>

png

plt.plot(x, y1, label='first')
plt.plot(x, y2, label='second')
plt.axis('off')
# plt.axis('off')Erase the axis with
plt.legend()

<matplotlib.legend.Legend at 0x7f05182f8e10>

png

21st

subplot

x = np.linspace(-3, 3, 10)
y1 = np.exp(x)
y2 = x*x

plt.subplot(1, 2, 1)
plt.plot(x, y1)

plt.subplot(1, 2, 2)
plt.plot(x, y2)

[<matplotlib.lines.Line2D at 0x7f0505d1d690>]

png

Object-oriented description

plt.figure object

fig = plt.figure()
type(fig)

matplotlib.figure.Figure




<Figure size 432x288 with 0 Axes>

fig = plt.figure()
ax1 = fig.add_subplot(1,2,1)
ax2 = fig.add_subplot(1,2,2)

ax1.plot(x, y1)
ax2.plot(x, y2)

[<matplotlib.lines.Line2D at 0x7f0505aead10>]

png

#Create a 1-by-2 plot Each axes object is returned as a list in axes
fig, axes = plt.subplots(nrows=1, ncols=2)
axes

array([<matplotlib.axes._subplots.AxesSubplot object at 0x7f0505bd2650>,
       <matplotlib.axes._subplots.AxesSubplot object at 0x7f0505a22a10>],
      dtype=object)

png

fig, axes = plt.subplots(nrows=1, ncols=2)
axes[0].plot(x, y1)
axes[1].plot(x, y2)

[<matplotlib.lines.Line2D at 0x7f0505894fd0>]

png

fig, axes = plt.subplots(nrows=3, ncols=3)
print(axes.shape)

(3, 3)

png

fig, axes = plt.subplots(nrows=3, ncols=3)
axes[1,2].plot(x,y2)

[<matplotlib.lines.Line2D at 0x7f05055cee90>]

png

fig, axes = plt.subplots(nrows=1, ncols=2)
axes[0].plot(x, y1, label='something')
axes[1].plot(x, y1)
axes[0].set_xlabel('xlabel1')
axes[0].set_ylabel('xlabel2')
axes[0].set_title('plot title')
axes[0].set_xticks([-3, -2, -1, 3])
axes[0].set_yticks([0, 10, 20])
axes[0].legend()
axes[1].axis('off')

(-3.3, 3.3, -0.9520004243731263, 21.08732441592866)

png

22nd

Adjust the size of the graph

x = np.linspace(-3, 3, 10)
y1 = np.exp(x)
y2 = np.exp(x)*2
fig, axes = plt.subplots()
axes.plot(x, y1)

[<matplotlib.lines.Line2D at 0x7f0505c7cb10>]

png

#In this case, a graph of 100 pixels x 100 pixels is displayed on the monitor.
fig, axes = plt.subplots(figsize=(1,1), dpi=100)
axes.plot(x, y1)

[<matplotlib.lines.Line2D at 0x7f0505308750>]

png

fig, axes = plt.subplots(figsize=(10, 3))
axes.plot(x, y1)

[<matplotlib.lines.Line2D at 0x7f05056641d0>]

png

Save as png

fig, axes = plt.subplots(2, 1, figsize=(10, 3))
axes[0].plot(x, y1, label='something')
axes[1].plot(x, x*x)

axes[0].set_title('first')
axes[1].set_title('second')

axes[0].set_xlabel('x')
axes[0].set_ylabel('y')
axes[1].set_xlabel('x')
axes[1].set_ylabel('y')

fig.savefig('savefig_sample.png')

png

fig, axes = plt.subplots(2, 1, figsize=(10, 3))
axes[0].plot(x, y1, label='something')
axes[1].plot(x, x*x)

axes[0].set_title('first')
axes[1].set_title('second')

axes[0].set_xlabel('x')
axes[0].set_ylabel('y')
axes[1].set_xlabel('x')
axes[1].set_ylabel('y')

plt.tight_layout() #Adjust the graph to make it easier to see

fig.savefig('savefig_sample.png')

png

fig, axes = plt.subplots()
axes.plot(x, y1, label='first')
axes.plot(x, y2, label='second')
axes.plot(x, y1+y2, label='first+second')
axes.legend()

<matplotlib.legend.Legend at 0x7f0505664f50>

png

Save as pdf

from matplotlib.backends.backend_pdf import PdfPages

pdf = PdfPages('savefig_sample.pdf')

from matplotlib.backends.backend_pdf import PdfPages

pdf = PdfPages('savefig_sample.pdf')

#------Graph creation-------
fig, axes = plt.subplots()
axes.plot(x, y1, label='first')
axes.plot(x, y2, label='second')
axes.plot(x, y1+y2, label='first+second')
axes.legend(loc=0)
#---------------------

#Save to pdf
pdf.savefig(fig)
#close processing (I will do it for the time being)
pdf.close()

png

Save a large number of graphs as pdf

pdf = PdfPages('savemultifig_sample.pdf')
for i in range(0, 10):
    #------Graph creation--------
    fig, axes = plt.subplots( )
    #It was designed so that the shape of the graph gradually changes. (Appropriate.)
    axes.plot(x, y1 + x*i)
    #Give it a title. Please confirm that you can search for characters in pdf.
    axes.set_title('ID:#{}'.format(i))
    #-----------------------

    #Save in for loop
    pdf.savefig(fig)

#Close after loop
pdf.close()

png

23rd

How to decorate the graph

import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

x = np.linspace(-3, 3, 10)
y = np.exp(x)
plt.plot(x, y)

[<matplotlib.lines.Line2D at 0x7f050476fd90>]

png

color: Graph line color ⇨ Specify a color name such as ‘red’ or ‘green’. You can use acronyms like ‘r’ and ‘g’.

lw (line width): Line thickness ⇨ Number. Please make it the size you like

ls (line style): Line type ⇨ Specify as ‘-’ or ‘–’. Either of these two is often used.

marker: Marker type ⇨ Specify as ‘o’ or ‘x’. The shape of the marker changes.

markersize: Marker size ⇨ Number. Please make it the size you like

markerfacecolor: Marker color ⇨ Specify the color name such as ‘red’ or ‘green’. You can use acronyms like ‘r’ and ‘g’.

markeredgecolor: Color in the marker frame ⇨ Specify the name of the color such as ‘red’ or ‘green’. You can use acronyms like ‘r’ and ‘g’.

markeredgewidth: Thickness of the marker frame ⇨ Number. Please make it the size you like

alpha: Plot transparency ⇨ Specify between 0 and 1 with float. The closer it is to 0, the higher the transparency.

plt.plot(x, y, color='red', lw=5, ls='--', marker='o', markersize=15, markerfacecolor='yellow', markeredgecolor='blue',
        markeredgewidth=4, alpha=0.5)

[<matplotlib.lines.Line2D at 0x7f05050fbd90>]

png

Scatter plot: plt.scatter ()

import pandas as pd
df = pd.read_csv('train.csv')
plt.scatter(df['Age'], df['Fare'], alpha=0.3)

<matplotlib.collections.PathCollection at 0x7f04b99fbf50>

png

Histogram: plt.hisgt ()

plt.hist(df['Age'])
plt.show()

/opt/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:829: RuntimeWarning: invalid value encountered in greater_equal
  keep = (tmp_a >= first_edge)
/opt/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:830: RuntimeWarning: invalid value encountered in less_equal
  keep &= (tmp_a <= last_edge)

png

plt.hist(df['Age'], bins=50)
plt.show()

png

Box plot: plt.boxplot ()

df = df.dropna(subset=['Age'])
plt.boxplot(df['Age'])
plt.show()