如何用ggplot可视化面板数据的加权百分比？

我有如下结构的面板数据，包含v1-v4的重复二元测量值，还有调查权重（custom_wt）和性别（gender）列。目标是：

• 计算并可视化任意年份/波次中v1取值为1的受访者的加权百分比，同理处理v2、v3、v4
• 横轴为v1/v2/v3/v4，纵轴为百分比

示例数据：

d <- data.frame(
  respondent_id = c(1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 6, 7, 7, 7, 8, 8, 8, 10),
  custom_wt = c(24, 24, 24, 26, 26, 26, 12, 12, 12, 14, 14, 14, 33, 33, 33, 32, 32, 32, 9, 9, 9, 10, 10, 10, 10),
  year = c(2002, 2004, 2006, 2002, 2004, 2006, 2002, 2004, 2006, 2002, 2004, 2006, 2002, 2004, 2006, 2002, 2004, 2006, 2002, 2004, 2006, 2002, 2004, 2006, 2006),
  v1 = c(0, 0, 1, NA, 1, 1, 0, NA, NA, 1, NA, NA, 1, 0, 0, 0, 0, 0, 1, 1, NA, 1, 0, 0, 0),
  v2 = c(0, 0, 1, NA, 1, 1, 0, NA, 1, NA, 1, 1, 0, NA, 1, 0, 1, 1, NA, 1, 1, 0, NA, 1, 1),
  v3 = c(0, 0, NA, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, NA, 0, 0, 1, 0, 0, 0, 0, 0, 0),
  v4 = c(NA, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, NA, 0, 0, 0, 1, 0, 0, NA, NA),
  gender = c(0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0)
)

解决步骤 #

1. 数据整理（宽转长格式） #

先把宽格式数据转为长格式，方便对v1-v4统一处理：

library(tidyverse)

d_long <- d %>%
  pivot_longer(cols = starts_with("v"), names_to = "variable", values_to = "value") %>%
  filter(!is.na(value))  # 剔除缺失值，若需将缺失值视为0，可替换为mutate(value = replace_na(value, 0))

2. 计算加权百分比 #

分两种场景计算：

场景1：不区分年份，计算每个变量的整体加权百分比

weighted_summary <- d_long %>%
  group_by(variable) %>%
  summarise(
    weighted_pct = sum(custom_wt[value == 1]) / sum(custom_wt) * 100,
    .groups = "drop"
  )

场景2：按年份分组，计算每个变量在各年份的加权百分比

weighted_summary_by_year <- d_long %>%
  group_by(year, variable) %>%
  summarise(
    weighted_pct = sum(custom_wt[value == 1]) / sum(custom_wt) * 100,
    .groups = "drop"
  )

3. 可视化 #

场景1：整体加权百分比柱状图

ggplot(weighted_summary, aes(x = variable, y = weighted_pct)) +
  geom_col(fill = "#2c3e50", width = 0.6) +
  geom_text(aes(label = sprintf("%.1f%%", weighted_pct)), vjust = -0.5) +
  labs(
    title = "各变量取值为1的加权百分比",
    x = "变量",
    y = "加权百分比 (%)"
  ) +
  theme_minimal() +
  ylim(0, max(weighted_summary$weighted_pct) + 10)  # 调整纵轴范围避免文字截断

场景2：按年份分组的分组柱状图

ggplot(weighted_summary_by_year, aes(x = variable, y = weighted_pct, fill = factor(year))) +
  geom_col(position = position_dodge(width = 0.7), width = 0.6) +
  geom_text(aes(label = sprintf("%.1f%%", weighted_pct)), 
            position = position_dodge(width = 0.7), vjust = -0.5) +
  labs(
    title = "各年份下变量取值为1的加权百分比",
    x = "变量",
    y = "加权百分比 (%)",
    fill = "年份"
  ) +
  theme_minimal() +
  ylim(0, max(weighted_summary_by_year$weighted_pct) + 10)

补充说明 #

计算逻辑：加权百分比 =（取值为1的受访者权重之和）/（该组所有受访者权重之和）×100，若需要将缺失的二元值视为0参与计算，只需修改数据整理步骤的缺失值处理方式即可。

内容的提问来源于stack exchange，提问作者a_todd12