森林图 (Forest Plot)¶

用途：Meta 分析、亚组分析结果的 OR/HR/MD 及置信区间的可视化展示。

交互式图表¶

悬停查看 HR 和 95% CI 数值，可缩放平移。

生成代码¶

RPythonSAS

library(plotly)

# 示例数据
forest_data <- data.frame(
  Subgroup = c("Overall", "Age < 65", "Age ≥ 65", "Male", "Female",
               "Region Asia", "Region EU", "Region US"),
  HR = c(0.72, 0.68, 0.78, 0.74, 0.70, 0.71, 0.73, 0.69),
  Lower = c(0.58, 0.48, 0.55, 0.53, 0.50, 0.49, 0.51, 0.47),
  Upper = c(0.89, 0.96, 1.10, 1.03, 0.98, 1.02, 1.04, 1.01)
)

fig <- plot_ly(forest_data) |>
  add_trace(
    x = ~HR, y = ~Subgroup,
    type = "scatter", mode = "markers",
    error_x = list(
      type = "data",
      symmetric = FALSE,
      arrayminus = ~HR - Lower,
      array = ~Upper - HR,
      color = "#1f77b4"
    ),
    marker = list(size = 10, color = "#1f77b4"),
    hovertemplate = "HR: %{x:.2f}<br>(%{customdata[0]:.2f}, %{customdata[1]:.2f})<extra>%{y}</extra>",
    customdata = ~stack(Lower, Upper)
  ) |>
  layout(
    title = "Forest Plot — Subgroup Analysis",
    xaxis = list(title = "Hazard Ratio (95% CI)", zeroline = TRUE, zerolinecolor = "gray"),
    yaxis = list(title = "", categoryorder = "array", categoryarray = rev(forest_data$Subgroup)),
    shapes = list(
      list(type = "line", x0 = 1, x1 = 1, y0 = -1, y1 = length(unique(forest_data$Subgroup)),
           line = list(dash = "dash", color = "gray", width = 1))
    ),
    margin = list(l = 100)
  )

fig

import plotly.graph_objects as go
import pandas as pd

forest_data = pd.DataFrame({
    "Subgroup": ["Overall", "Age < 65", "Age ≥ 65", "Male", "Female",
                 "Region Asia", "Region EU", "Region US"],
    "HR": [0.72, 0.68, 0.78, 0.74, 0.70, 0.71, 0.73, 0.69],
    "Lower": [0.58, 0.48, 0.55, 0.53, 0.50, 0.49, 0.51, 0.47],
    "Upper": [0.89, 0.96, 1.10, 1.03, 0.98, 1.02, 1.04, 1.01],
})

fig = go.Figure()

fig.add_trace(go.Scatter(
    x=forest_data["HR"],
    y=forest_data["Subgroup"],
    mode="markers+lines",
    marker=dict(size=10, color="#1f77b4"),
    error_x=dict(
        type="data",
        symmetric=False,
        arrayminus=forest_data["HR"] - forest_data["Lower"],
        array=forest_data["Upper"] - forest_data["HR"],
        color="#1f77b4"
    ),
    hovertemplate="HR: %{x:.2f}<br>(%{customdata[0]:.2f}, %{customdata[1]:.2f})<br>%{y}<extra></extra>",
    customdata=forest_data[["Lower", "Upper"]]
))

fig.add_vline(x=1, line_dash="dash", line_color="gray")

fig.update_layout(
    title="Forest Plot — Subgroup Analysis",
    xaxis_title="Hazard Ratio (95% CI)",
    yaxis=dict(categoryorder="array", categoryarray=forest_data["Subgroup"][::-1]),
    height=400,
    margin=dict(l=120)
)

fig.show()

proc sgplot data=forest_data;
    highlow y=subgroup low=lower high=upper / 
        type=bar lineattrs=(color=navy) barwidth=0.2;
    scatter y=subgroup x=hr / 
        markerattrs=(symbol=squarefilled color=navy size=10);
    refline 1 / axis=x lineattrs=(pattern=dash color=gray);
    xaxis label="Hazard Ratio (95% CI)";
run;

关键参数¶

参数	说明
`error_x`	置信区间范围
`xref = 1`	参考线位置 (OR/HR=1, MD=0)
`categoryorder`	亚组排序（通常倒序）

AI Prompt 模板¶

请用 Plotly 生成一张森林图。数据为临床试验亚组分析的 HR 和 95% CI，
包含 Overall 和亚组变量。要求：HR 以 1 为参考线，亚组展示在 y 轴，
水平排列，右侧留出空间显示 HR 值和 p 值。

参考方法¶