CODEWITHSUNDEEP
pythonmatplotlib
Naz
Naz

Reputation: 495

Set 'y' axis to scientific notation

I want my y axis to be formatted in scientific notation.

I have tried the matplotlib documentation, but it ignores my command.

import matplotlib.pyplot as plt
import numpy as np


x = np.random.randint(1e4, size=200)
y = np.random.randint(1e4, size=200)

plt.ticklabel_format(axis='both', style='sci')
plt.xlabel('x')
plt.ylabel('y')
plt.scatter(x,y, color='b', s=5, marker=".")

plt.show()

My output just appears in none-scientific notation.

For example I expect the ylabel 1000 to be 1E03.

This code is just an example.

I have a sub-plot where plot 1 and 3 are in scientific notation, but plot 2 is in non-scientific notation.

Upvotes: 17

Views: 43624

Answers (3)

Lukas
Lukas

Reputation: 163

The other solutions didn't work for me (after the plot command) on Matplotlib 3.4.3, but what worked was (here for the y-axis):

# set the y axis ticks to 10^x
ax.yaxis.set_major_formatter(mpl.ticker.FuncFormatter(lambda x, _: '{:.0e}'.format(x)))
# set the y axis tick labels to 10^x
ax.set_yticklabels(['$10^{'+str(int(np.log10(y)))+'}$' for y in ax.get_yticks()])

Upvotes: 0

Meysam Sadeghi
Meysam Sadeghi

Reputation: 1603

You should just add scilimits=(4,4) to your command

plt.ticklabel_format(axis='both', style='sci', scilimits=(4,4))

for example your code will become:

x = np.random.randint(1e4,size=200)
y = np.random.randint(1e4,size=200)
plt.ticklabel_format(axis='both', style='sci', scilimits=(4,4))
plt.xlabel('x')
plt.ylabel('y')
plt.scatter(x,y, color='b', s=5, marker=".")
plt.show()

Upvotes: 16

kklocker
kklocker

Reputation: 166

Passing

plt.ticklabel_format(axis='both', style='sci', scilimits=(0,0))

worked for me.

(as @Meysam-Sadeghi suggested, but without plt.subplots())

Upvotes: 9

Related Questions