Python-matplotlib | Draw dual y-axis graphics (legend settings)

Release： 2023-08-09 16:06:30

forward

2385 people have browsed it

Today I will introduce to you how to use Python’s matplotlib library to draw a double y-axis graph And the legend setting problem , I hope it will be helpful to everyone. If you have any questions or suggestions, you can send a private message to the editor.

Rendering preview:

Python-matplotlib | Draw dual y-axis graphics (legend settings)

Sample data:

df = pd.read_csv(&#39;jobdata.csv&#39;)

Copy after login

Python-matplotlib | Draw dual y-axis graphics (legend settings)

# 1. Double y-axis line chart

##1 . Position number line chart

##
colors = ["#51C1C8", "#536D84","#E96279"] plt.figure(figsize=(16, 8)) ax1 = plt.subplot(111) ax1.set_ylim(0,1200) lin0 = ax1.plot(x_data, y_data1, marker='o', color=colors[0], label='岗位数量') for x, y in enumerate(y_data1): plt.text(x - 0.2, y+5, y) ax1.set_ylabel('岗位数量',fontsize=12) plt.legend() plt.title("各城市Java岗位数量") plt.show()
Copy after login

##2. Add a y-axis through ax1.twinx():

Python-matplotlib | Draw dual y-axis graphics (legend settings)

# 增加y轴
ax2 = ax1.twinx()

ax2.set_ylim(0,60)
lin1 = ax2.plot(x_data, y_data2, linestyle=&#39;--&#39;, marker=&#39;o&#39;, c=colors[1], label=&#39;平均最低薪资&#39;) 
for x, y in enumerate(y_data2):
    plt.text(x - 0.1, y+1, y)
lin2 = ax2.plot(x_data, y_data3, linestyle=&#39;--&#39;, marker=&#39;o&#39;, c=colors[2], label=&#39;平均最高薪资&#39;)
for x, y in enumerate(y_data3):
    plt.text(x - 0.1, y+1, y)
ax2.set_ylabel(&#39;平均薪资(万/年)&#39;,fontsize=12)
plt.legend()
plt.title("各城市Java岗位数量和薪资水平状况")
plt.show()

Copy after login

重点：细心的小伙伴可能发现了图没有问题，但是右上角的图例只显示了平均最低薪资和平均最高薪资，但是岗位数量的图例并没有显示。

3. 单独设置图例

ax1.legend(loc=&#39;best&#39;)
ax2.legend(loc=&#39;best&#39;)

Copy after login

Python-matplotlib | Draw dual y-axis graphics (legend settings)

看着感觉没什么变化，实际上仔细看会发现平均最低薪资、平均最高薪资、岗位数量三个图例都显示出来了，只不过岗位数量图例被盖住了，我们可以移动一下位置看看：

ax1.legend(loc=2)
ax2.legend(loc=1)

Copy after login

Python-matplotlib | Draw dual y-axis graphics (legend settings)

这样看就比较直观了，但是我就想把三个图例放一起不可以吗？

当然可以!

3. 设置组合图例

lines = lin0+lin1+lin2
labs = [label.get_label() for label in lines]
plt.legend(lines,labs)

Copy after login

Python-matplotlib | Draw dual y-axis graphics (legend settings)

大功告成！

但是！如果是柱状图＋折线图的情况，效果还一样吗？

2、双y轴柱状图+折线图

1. 修改岗位数量为柱状图

plt.figure(figsize=(16, 8))
a1 = plt.subplot(111)
a1.set_ylim(0,1200)
bar = a1.bar(x_data, y_data1, color=colors[0], label=&#39;岗位数量&#39;) 
for x, y in enumerate(y_data1):
    plt.text(x - 0.2, y+5, y)
a1.set_ylabel(&#39;岗位数量&#39;,fontsize=12)

...

lines = bar+lin1+lin2
labs = [label.get_label() for label in lines]
plt.legend(lines,labs)

Copy after login

直接报错了！ Python-matplotlib | Draw dual y-axis graphics (legend settings)

Python-matplotlib | Draw dual y-axis graphics (legend settings)

The prompt type is inconsistent. It is obviously a problem with the type of bar and line. Let’s check the source code:

matplotlib.axes.Axes.plot:

Python-matplotlib | Draw dual y-axis graphics (legend settings)

matplotlib.axes.Axes.bar：

Python-matplotlib | Draw dual y-axis graphics (legend settings)

ax.plot returns a Line2D type list, ax.bar returns a patches Type tuple.

#After finding the root cause, we can just make a combination of line2D and patches.

2. 设置Line2D和patches的组合图例

legend_handles = [ 
    Line2D([], [], linewidth=1, ls=&#39;--&#39;, lw=2, c=colors[2], label=&#39;平均最高薪资&#39;),
    Line2D([], [], linewidth=1, lw=2, c=colors[1], label=&#39;平均最低薪资&#39;),
    patches.Rectangle((0, 0), 1, 1, facecolor=colors[0],label=&#39;岗位数量&#39;)
]
plt.legend(handles=legend_handles, loc=&#39;best&#39;, fontsize=14)

Copy after login

效果：

其他参数大家可以自行尝试修改，对比前后效果，加深理解。

The above is the detailed content of Python-matplotlib | Draw dual y-axis graphics (legend settings). For more information, please follow other related articles on the PHP Chinese website!