Matplotlib如何为多变量依赖量在单个图例条目合并多条曲线

• 计算同时依赖两个变量v1和v2的两个量q1和q2。
• 选取若干个v1的取值，绘制两个量随v2变化的曲线。
• 每个选中的v1取值对应单个图例条目，条目内需同时展示v1取值、q1和q2的均值及误差范围。

我此前参考了单条线+填充误差带的图例实现方案，但无法适配当前更复杂的多信息整合需求。

核心需求为实现上述规则的自定义图例，以下是测试代码：

import numpy as np
import matplotlib.pyplot as plt
from matplotlib import cm

def calc_quant1(x,y):
    q1 = x + y
    q1err = np.sqrt(x+y)
    return q1,q1err

def calc_quant2(x,y):
    q2 = x - y
    q2err = np.sqrt((x+y)/2)
    return q2,q2err


# variables
var1 = np.linspace(-6,10,5)
var2 = np.linspace(10,20,21)

# resulting quantities (preallocation)
quant1 = np.zeros((len(var1),len(var2)))
quant2 = np.zeros((len(var1),len(var2)))
quant1err = np.zeros((len(var1),len(var2)))
quant2err = np.zeros((len(var1),len(var2)))
dy_dx = np.zeros(len(var2))

# plot colors
colors = cm.jet(np.linspace(0,1,len(var1)))

# reference value
xref = 12

# initialize figure
fig = plt.figure(figsize=(8,6))
fig.suptitle('Title')
ax = fig.add_subplot(111)
ax.axvline(x=xref,color='black',linestyle='dashed',label='$x={}$'.format(xref))
    
# compute and plot
for i in np.arange(len(var1)):
    
    # select "measured" voltage
    v1 = var1[i] * np.ones(len(var2))
    # 注：原代码未定义var1err变量，运行时需补充定义
    # v1err = var1err[i] * np.ones(len(var2))
    
    # compute heat flux
    quant1[i,:], quant1err[i,:] = calc_quant1(v1,var2)
    quant2[i,:], quant2err[i,:] = calc_quant2(v1,var2)
    
    # compute gradient near reference point (suboptimal)
    x1 = var2[var2<xref][-1]
    x2 = var2[var2>xref][0]
    y1 = quant2[i][var2<xref][-1]
    y2 = quant2[i][var2>xref][0]
    dy_dx[i] = (y2-y1)/(x2-x1)
    
    # plot results
    label1 = r'$v_{{1}}={:.0f}$ [unit]'.format(var1[i])
    label2 = r'$q_{{1}}(10)={:.0f}$ [unit]'.format(quant1[i,0])
    ax.plot(var2,quant1[i,:], linestyle='dotted', linewidth=2, color=colors[i])
    ax.plot(var2,quant2[i,:], linestyle='solid', linewidth=1, color=colors[i])
    ax.fill_between(var2,quant1[i,:]-quant1err[i,:],quant1[i,:]+quant1err[i,:], alpha=0.16, color=colors[i])
    ax.fill_between(var2,quant2[i,:]-quant2err[i,:],quant2[i,:]+quant2err[i,:], alpha=0.16, color=colors[i],
                    label=label1+'\n'+label2)


# finalize figure
ax.set_xlim(np.min(var2),np.max(var2))
ax.set_xlabel('$v_{2}$ [unit]')
ax.set_ylabel('$q_{2}$ [unit]')
plt.tight_layout()
ax.annotate(text=r'$Q_{1}=v_{1}+v_{2}$',
            xy=(0.50,0.77), xycoords='axes fraction',
            xytext=(1.05,0.95), textcoords='axes fraction',
            arrowprops=dict(arrowstyle='->',connectionstyle='arc3,rad=0.2'),
            fontsize=20)
ax.annotate(text=r'$Q_{2}=v_{1}-v_{2}$',
            xy=(0.70,0.35), xycoords='axes fraction',
            xytext=(1.05,0.85), textcoords='axes fraction',
            arrowprops=dict(arrowstyle='->',connectionstyle='arc3,rad=0.1'),
            fontsize=20)
leg = ax.legend(loc='upper right', bbox_to_anchor=(1.475,0.775), labelspacing=1)
ax.grid()
plt.show()

内容的提问来源于stack exchange，提问作者ejok