如何在Plotly中为散点图的多个trace绘制趋势线

方案1：基于现有graph_objects写法手动添加 #

思路：对每个散点序列单独做线性拟合，将拟合结果作为线图trace加入画布即可，需用到numpy做数值计算：

import plotly.graph_objects as go
import numpy as np

fig = go.Figure()

# 添加道格拉斯冷杉原始散点
fig.add_trace(go.Scatter(x=df_df['Circumference (meters)'], 
                         y=df_df['Height (meters)'], 
                         name='Douglas Fir', mode='markers')
             )
# 计算道格拉斯冷杉线性趋势线参数
z1 = np.polyfit(df_df['Circumference (meters)'], df_df['Height (meters)'], 1)
p1 = np.poly1d(z1)
# 添加道格拉斯冷杉趋势线
fig.add_trace(go.Scatter(x=df_df['Circumference (meters)'], 
                         y=p1(df_df['Circumference (meters)']), 
                         name='Douglas Fir 趋势线', 
                         mode='lines',
                         line=dict(dash='dash'))
             )

# 添加白松原始散点
fig.add_trace(go.Scatter(x=df_wp['Circumference (meters)'], 
                         y=df_wp['Height (meters)'],  
                         name='White Pine',mode='markers')
             )
# 计算白松线性趋势线参数
z2 = np.polyfit(df_wp['Circumference (meters)'], df_wp['Height (meters)'], 1)
p2 = np.poly1d(z2)
# 添加白松趋势线
fig.add_trace(go.Scatter(x=df_wp['Circumference (meters)'], 
                         y=p2(df_wp['Circumference (meters)']), 
                         name='White Pine 趋势线', 
                         mode='lines',
                         line=dict(dash='dash'))
             )

fig.update_layout(title="Tree Circumference vs Height (meters)",
                  xaxis_title=df_df['Circumference (meters)'].name,
                  yaxis_title=df_df['Height (meters)'].name,
                  title_x=0.5)

fig.show()

方案2：用plotly.express简化实现（推荐） #

思路：将两个数据集合并为带分类标签的DataFrame，直接调用px内置的趋势线参数自动生成，无需手动计算拟合过程：

import plotly.express as px
import pandas as pd

# 给两个数据集添加树种分类列
df_df['树种'] = 'Douglas Fir'
df_wp['树种'] = 'White Pine'
# 合并为全量数据集
df_all = pd.concat([df_df, df_wp])

# 绘制散点图，自动按树种分类，添加OLS线性趋势线
fig = px.scatter(df_all, 
                 x='Circumference (meters)', 
                 y='Height (meters)', 
                 color='树种',
                 trendline='ols', # 指定用普通最小二乘法生成线性趋势线
                 title='Tree Circumference vs Height (meters)')
fig.update_layout(title_x=0.5)
fig.show()

实现效果参考 #

内容的提问来源于stack exchange，提问作者David 54321