如何在ggplot2柱状图中添加置信区间的阴影渐变效果？

我想复现一种比常规误差须更直观的置信区间视觉效果（带阴影/渐变），尝试用geom_linerange和geom_errorbar实现但没达到预期，附上我的数据和代码寻求帮助：

my_df <- 
tibble::tribble(~response, ~estimate, ~lower_ci, ~upper_ci,
                "little_bit", 0.353477, 0.255625, 0.451747,
                "no", 0.307639, 0.250436, 0.375393,
                "very", 0.338883, 0.301007, 0.37572310)


ggplot(my_df, aes(x = reorder(response, -estimate), y = estimate)) +
  geom_linerange(aes( ymin = lower_ci, ymax = upper_ci),
                 width = 0.9, 
                 size = 45, 
                 color = "red",
                 alpha  = 0.7
                ) +
  geom_errorbar(aes(ymin = estimate, ymax = estimate),
                width = 0.9, 
                size = 2 ,
                color = "#6EB3FF") 

要实现带渐变效果的置信区间可视化，推荐两种实用方法：

方法一：基于ggplot2原生函数实现渐变阴影 #

将离散x轴转换为数值并设置边界，用geom_ribbon绘制分上下段的渐变置信区间，再叠加估计值线条：

library(ggplot2)
library(tibble)

my_df <- 
  tibble::tribble(~response, ~estimate, ~lower_ci, ~upper_ci,
                  "little_bit", 0.353477, 0.255625, 0.451747,
                  "no", 0.307639, 0.250436, 0.375393,
                  "very", 0.338883, 0.301007, 0.37572310)

# 为离散类别分配数值及左右边界
my_df <- my_df %>%
  mutate(x_num = as.numeric(reorder(response, -estimate)),
         x_min = x_num - 0.4,
         x_max = x_num + 0.4)

ggplot(my_df) +
  # 置信区间下半段渐变阴影
  geom_ribbon(aes(x = x_num, ymin = lower_ci, ymax = estimate, xmin = x_min, xmax = x_max),
              fill = "#FF6B6B", alpha = 0.3) +
  # 置信区间上半段渐变阴影
  geom_ribbon(aes(x = x_num, ymin = estimate, ymax = upper_ci, xmin = x_min, xmax = x_max),
              fill = "#FF6B6B", alpha = 0.3) +
  # 叠加估计值粗线条
  geom_segment(aes(x = x_min, xend = x_max, y = estimate, yend = estimate),
               color = "#6EB3FF", size = 2) +
  # 还原x轴类别标签
  scale_x_continuous(breaks = my_df$x_num, labels = my_df$response) +
  labs(x = "response", y = "estimate") +
  theme_minimal()

方法二：用ggdist包快速生成专业渐变区间 #

ggdist专为分布与区间可视化设计，一行代码即可实现带渐变的区间条：

library(ggplot2)
library(tibble)
library(ggdist)

my_df <- 
  tibble::tribble(~response, ~estimate, ~lower_ci, ~upper_ci,
                  "little_bit", 0.353477, 0.255625, 0.451747,
                  "no", 0.307639, 0.250436, 0.375393,
                  "very", 0.338883, 0.301007, 0.37572310)

ggplot(my_df, aes(y = reorder(response, -estimate), x = estimate, xmin = lower_ci, xmax = upper_ci)) +
  # 渐变区间条，size控制宽度，fill设置渐变色调
  geom_slinterval(fill = "#FF6B6B", alpha = 0.7, size = 15) +
  # 叠加估计值标记点
  geom_point(aes(x = estimate), color = "#6EB3FF", size = 5) +
  labs(y = "response", x = "estimate") +
  theme_minimal()

说明：方法一灵活性强，可自定义渐变方向、透明度等细节；方法二更简洁高效，适合快速生成符合专业标准的区间可视化效果。

内容的提问来源于stack exchange，提问作者user3148607