9

I have a horizontal barplot, for example, a simplified version of the example from the seaborn documentation:

import seaborn as sns
import matplotlib.pyplot as plt

f, ax = plt.subplots(figsize=(6, 15))

crashes = sns.load_dataset("car_crashes").sort_values("total", ascending=False)

sns.barplot(x="total", y="abbrev", data=crashes,
            label="Total", color="b")

ax.set(xlim=(0, 24), ylabel="",
       xlabel="Automobile collisions per billion miles")

plt.show()

How can I get the bars labeled with the value for each bar?

I tried this approach for vertical bars (How to add percentages on top of bars in seaborn), but it doesn't seem to work. Changing height to width doesn't have the effect I assumed it would.

for p in ax.patches:
    height = p.get_width()
    ax.text(p.get_y()+p.get_height()/2.,
            height + 3,
            '{:1.2f}'.format(height),
            ha="center")

I'm assuming the horizontal plot works differently?

tdy
  • 36,675
  • 19
  • 86
  • 83
max
  • 4,141
  • 5
  • 26
  • 55
  • 1
    the `text` requires the first two argument as `x` and `y`, i.e. `ax.text(x,y)`. You re providing it with `ax.text(y,x)`. I would also suggest not to label the `width` by `"height"` as this will confuse everyone including yourself. – ImportanceOfBeingErnest Apr 13 '18 at 15:39

3 Answers3

8

Got it, thanks to @ImportanceOfBeingErnest

This worked for me

for p in ax.patches:
    width = p.get_width()    # get bar length
    ax.text(width + 1,       # set the text at 1 unit right of the bar
            p.get_y() + p.get_height() / 2, # get Y coordinate + X coordinate / 2
            '{:1.2f}'.format(width), # set variable to display, 2 decimals
            ha = 'left',   # horizontal alignment
            va = 'center')  # vertical alignment
Mr. T
  • 11,960
  • 10
  • 32
  • 54
max
  • 4,141
  • 5
  • 26
  • 55
7

As of matplotlib 3.4.0

Use the new built-in ax.bar_label, which will automatically label bar containers regardless of orientation:

fig, ax = plt.subplots(figsize=(6, 8))
sns.barplot(x="total", y="abbrev", data=crashes)

# new helper method to auto-label bars
ax.bar_label(ax.containers[0])

If the bars are grouped by hue, call ax.bar_label on all the containers:

fig, ax = plt.subplots(figsize=(5, 6))
ax = sns.barplot(x="tip", y="day", hue="smoker", data=tips)

# grouped bars will have multiple containers
for container in ax.containers:
    ax.bar_label(container)

tdy
  • 36,675
  • 19
  • 86
  • 83
0

Thank you very much for this. It helped me a lot, but i ran to a problem, where percents had to many digits after decimal point, the format can be then simply specified:

for container in ax.containers:
    ax.bar_label(container,size=8,fmt='%.1f')
Richard Dvořák
  • 319
  • 2
  • 3