主页 / coding / 问题 / 79207124
Accepted
MuhammedYunus
Asked: 2024-11-20 20:04:12 +0800 CST

如何使用 plt.bar_label() 有选择地标记条形图？

  • 772

matplotlib当使用绘制条形图时bars = plt.bar(...)，可以将返回的对象传递plt.bar_label(bars, labels)给以便于标记。

plt.bar_label(bars, labels)标记所有条形。我只想标记一个子集（前三个），最好使用plt.bar_label()。我不想调用 plt.text()或plt.annotate()。

有没有办法从中切出我想要的条形图bars，并且只使用标签plt.bar_label()？

下面的例子显示了所有条形图都已贴上标签的默认行为。

在此处输入图片描述

#Imports
import matplotlib.pyplot as plt
import pandas as pd

# Data for testing
values = [10, 9, 10.5, 4, 3.1, 2]
labels = [
    'sensor calibration', 'sensor response', 'noise floor estimate',
    'noise floor theta', 'weighted response coef', 'linear estimate'
]

#Bar plot with labels
ax = plt.figure(figsize=(5, 2)).add_subplot()
bars = ax.bar(range(6), values, color='lightgray')
ax.set(xlabel='x', ylabel='value', title='Data')
ax.spines[['top', 'right']].set_visible(False)

#Add all bar labels
label_kwargs = dict(rotation=90, weight='bold', fontsize=8, label_type='center')
ax.bar_label(bars, labels, **label_kwargs)
matplotlib

2 个回答

  1. 您可以切出所需的条形，但随后必须将它们重新封装在 中BarContainer。生成的对象可以plt.bar_label()像以前一样传递给 。

    在此处输入图片描述

    #bars = <from OP's code>

#Only label the first k bars
from matplotlib.container import BarContainer
k = 3

# Slice out the required bars & their values, and
#   re-encapsulate in a `BarContainer`
bars_k = BarContainer(bars[:k], datavalues=bars.datavalues[:k])
labels_k = labels[:k]

#Use bar_label() as before
ax.bar_label(bars_k, labels_k, **label_kwargs)
    • 1
  2. Best Answer

    一个简单的选择就是为不应该标记的条形提供空标签：

    import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

# Data for testing
values = [10, 9, 10.5, 4, 3.1, 2]
labels = [
    'sensor calibration', 'sensor response', 'noise floor estimate',
    'noise floor theta', 'weighted response coef', 'linear estimate'
]

# Bar plot with labels
fig, ax = plt.subplots(figsize=(5, 3))
bars = ax.bar(range(6), values, color='lightgray')
ax.set(xlabel='x', ylabel='value', title='Data')
ax.spines[['top', 'right']].set_visible(False)

# Add centered bar labels
k = 3
label_kwargs = dict(rotation=90, weight='bold', fontsize=8, label_type='center')
ax.bar_label(bars, labels[:k] + [''] * (len(bars) - k), **label_kwargs)

plt.show()

    选定条形的 bar_label

    以下变体将长条的标签置于中央，将短条的标签置于顶部：

    # Bar plot with labels
fig, ax = plt.subplots(figsize=(5, 3))
bars = ax.bar(range(6), values, color='lightgray')
ax.set(xlabel='x', ylabel='value', title='Data')
ax.spines[['top', 'right']].set_visible(False)

cutoff = 5  # bars higher will get centered labels, lower bars get labels on top
# Add the centered bar labels
label_kwargs = dict(rotation=90, weight='bold', fontsize=8, label_type='center')
ax.bar_label(bars, ['' if val <= cutoff else lbl for val, lbl in zip(values, labels)], **label_kwargs)
# Add bar labels at the top of the other bars
label_kwargs['label_type'] = 'edge'
label_kwargs['padding'] = 2
ax.bar_label(bars, ['' if val > cutoff else lbl for val, lbl in zip(values, labels)], **label_kwargs)

plt.show()

    bar_label 居中还是位于顶部取决于条形高度

    • 1

相关问题