Why is my matplotlib bar chart compressing the x-axis when I do 'log'

Question

I am trying to make a bar chart with matplotlib for python. I can make a normal chart, but when I put it in logarithmic mode so that I can see the data better, the x-axis looks as if it is compressing. What is going on? Note: all data records have a minimum value of 1.

    x = [t[0] for t in data]
    y = [t[1]  for t in data]

    x_pos = np.arange(len(x))

    plt.bar(x_pos, y, color='blue', log=False)
    plt.xlabel(x_label)
    plt.ylabel(y_label)
    plt.title(title)
    plt.xlim([0, len(x)])
    #plt.yscale('log')
    #plt.semilogy(x_pos, np.exp(-x_pos/5.0))
    plt.savefig(output_path + '/' + filename)

Yields:

But just by changing log=False to log=True I get:

What am I doing wrong? I just want to get a compressed view of the first graph on the y axis. As you can see, I also tried yscale('log') but I get the same result.

Thanks!

EDIT: Looks like it has something to do with the previous lines, when I remove the first line it works fine, but is unsorted:

data = sorted(data, key=lambda tup: tup[1], reverse=True)

    # tuples (pair,count)
    x = [t[0] for t in data]
    y = [t[1]  for t in data]

score 3 · Accepted Answer · 2016-07-08 19:36:13Z

3

If you just want a log scale on the y axis and not the x use:

import matplotlib.pyplot as plt

plt.semilogy(x,y) #plots log on y axis

plt.semilogx(x,y) #plots log on x axis

Alternatively, you can just set the axis' to log:

ax = plt.subplot(224)

ax.set_xscale("log")
ax.set_yscale("log")

plt.bar(x,y)

edited Jul 8, 2016 at 19:36

answered Jul 8, 2016 at 19:30

user6530460

Sign up to request clarification or add additional context in comments.

2 Comments

skaz Over a year ago

I have the same problem. It still compresses on the x axis, just like in my example.

skaz Over a year ago

I tried both methods but I get the same result. The X axis is compressed, like in my second picture.

Maxwell Collard · Accepted Answer · 2016-07-08 20:04:33Z

1

It isn't compressing your x-axis; the minimum value of your second plot's y-axis is 1 (10^0), which it appears is the height of the smallest set of bars. Hence, the rightmost bars are off the y scale on your semilog plot.

Try adding, e.g.,

plt.ylim([0.1, 100])

to enforce the visibility of the rightmost bars. (A log-scale axis can't have a minimum limit of 0, for obvious reasons.)

answered Jul 8, 2016 at 20:04

Maxwell Collard

2,8473 gold badges18 silver badges12 bronze badges

1 Comment

skaz Over a year ago

Thank you, sir! That was indeed the problem!

Collectives™ on Stack Overflow

Why is my matplotlib bar chart compressing the x-axis when I do 'log'

2 Answers 2

2 Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related