瀑布图 (Waterfall Plot)¶

用途：展示每个受试者靶病灶较基线的最佳变化百分比，常用于肿瘤学疗效评估。

交互式图表¶

每个条形代表一位受试者，绿色虚线 = PR 阈值（-30%），红色虚线 = PD 阈值（+20%）。悬停查看具体变化值。

生成代码¶

RPython

library(plotly)

wf_data <- data.frame(
  Subject = paste0("S", 1:40),
  PctChange = c(-100, -85, -72, -65, -60, -55, -50, -45, -40, -35,
                -30, -28, -25, -22, -20, -18, -15, -12, -10, -8,
                -5, -2, 0, 5, 8, 10, 12, 15, 18, 20,
                22, 25, 28, 30, 35, 40, 50, 60, 75, 100)
)

wf_data$color <- ifelse(wf_data$PctChange <= -30, "#1f77b4",  # 缓解
                        ifelse(wf_data$PctChange <= 20, "#ff7f0e",  # 稳定
                               "#d62728"))  # 进展

fig <- plot_ly(wf_data, x = ~Subject, y = ~PctChange,
               type = "bar", marker = list(color = ~color),
               hovertemplate = "Subject: %{x}<br>Change: %{y:.1f}%<extra></extra>")

fig <- fig %>% layout(
  title = "Waterfall Plot — Best Tumor Response",
  xaxis = list(title = "Subject", showticklabels = FALSE),
  yaxis = list(title = "Best Change from Baseline (%)"),
  shapes = list(
    list(type = "line", x0 = -1, x1 = 40, y0 = -30, y1 = -30,
         line = list(dash = "dash", color = "green")),
    list(type = "line", x0 = -1, x1 = 40, y0 = 20, y1 = 20,
         line = list(dash = "dash", color = "red"))
  )
)
fig

import plotly.graph_objects as go
import pandas as pd
import numpy as np

wf_data = pd.DataFrame({
    "Subject": [f"S{i}" for i in range(1, 41)],
    "PctChange": [-100,-85,-72,-65,-60,-55,-50,-45,-40,-35,
                  -30,-28,-25,-22,-20,-18,-15,-12,-10,-8,
                  -5,-2,0,5,8,10,12,15,18,20,
                  22,25,28,30,35,40,50,60,75,100]
})

colors = ["#1f77b4" if x <= -30 else "#ff7f0e" if x <= 20 else "#d62728" 
          for x in wf_data["PctChange"]]

fig = go.Figure(data=go.Bar(
    x=wf_data["Subject"], y=wf_data["PctChange"],
    marker_color=colors,
    hovertemplate="Subject: %{x}<br>Change: %{y:.1f}%<extra></extra>"
))

fig.add_hline(y=-30, line_dash="dash", line_color="green")
fig.add_hline(y=20, line_dash="dash", line_color="red")

fig.update_layout(
    title="Waterfall Plot — Best Tumor Response",
    xaxis_showticklabels=False,
    yaxis_title="Best Change from Baseline (%)"
)
fig.show()

关键参数¶

y = -30%: 客观缓解（RECIST 1.1 PR 阈值）
y = 20%: 疾病进展（RECIST 1.1 PD 阈值）

AI Prompt 模板¶

请用 Plotly 生成瀑布图，展示受试者靶病灶最佳变化百分比。
数据包含 subject_id 和 pct_change 两列。
按变化从小到大排序，在 -30% 和 20% 处画参考线。
<-30% 蓝色（缓解），-30%~20% 橙色（稳定），>20% 红色（进展）。

参考方法¶