如何根据 Y 轴“高度”为每个条添加不同的 alpha

Question

我想将 alpha 添加到我的条形图 (ggplot2) 但每个条我想要一个不同的 alpha，它应该取决于 y 轴高度。 我对 r 不太有经验，所以我还在苦苦挣扎:) 到目前为止，我的代码是这样的。 正确的选择是二项分布，并且只标注了 1 或 0。

#Data preparation#
DataMerged235$treatment <- factor(DataMerged235$treatment, levels = c("odour & flavour", "only odour", "only flavour"))

#Figure#
 fig235 <- ggplot (DataMerged235, aes(x = treatment, y = correct_final)) +
  stat_summary(fun.y = "mean", geom = "bar", fill="dodgerblue4") +
  stat_summary(fun.data = "mean_cl_boot", geom="errorbar", width = 0.2) +
  ylab("Proportion of correct choices") + xlab("") + theme_bw(16) +
  geom_abline(intercept = 0.5, slope = 0, color = "black", linetype = 2) +
  coord_cartesian(ylim = c(0.4, 1.15))

 print(fig235)

这里是当前输出：https://ibb.co/mHVSwpt

我想补充一点，条形的 y max 越高意味着 alpha 越暗。在我们的示例中，这将导致 alpha 降低（从左到右）。

我数据中的随机部分：

structure(list(colony = c(14, 5, 12, 11, 12, 14, 12, 18, 12, 
12), correct_initial = c(1, 0, 1, 1, 1, 1, 0, 0, 1, 1), correct_final = c(1, 
0, 1, 1, 1, 1, 0, 0, 1, 1), treatment = c("odour & flavour", 
"only odour", "odour & flavour", "only odour", "odour & flavour", 
"odour & flavour", "odour & flavour", "only flavour", "only flavour", 
"odour & flavour")), row.names = c(NA, -10L), class = c("tbl_df", 
"tbl", "data.frame"))

Answer 1

我还没有弄清楚如何在 stat_summary 行中执行此操作，但另一种方法是预先计算您要使用的统计信息。

library(dplyr)
mtcars %>%
    group_by(gear) %>%
    summarize(mean = mean(mpg),
              stats = mean_cl_boot(mpg), .groups = "drop") -> mtcars_stats
> mtcars_stats
# A tibble: 3 x 3
   gear  mean stats$y $ymin $ymax
  <dbl> <dbl>   <dbl> <dbl> <dbl>
1     3  16.1    16.1  14.4  17.7
2     4  24.5    24.5  21.8  27.4
3     5  21.4    21.4  16.3  26.6

那么就很简单了：

ggplot(aes(gear, mean, ymin = stats$ymin, ymax = stats$ymax)) +
  geom_col(fill = "dodgerblue4", aes(alpha = mean)) +
  geom_errorbar(width = 0.2)

编辑 - 使用来自 OP 的新示例数据进行更新：

DataMerged235 %>%
    mutate(treatment = factor(treatment, levels = c("odour & flavour", "only odour", "only flavour"))) %>%
    group_by(treatment) %>%
    summarize(mean = mean(correct_final),
              stats = mean_cl_boot(correct_final), .groups = "drop") %>%
    ggplot(aes(treatment, mean, ymin = stats$ymin, ymax = stats$ymax)) +
    geom_col(fill = "dodgerblue4", aes(alpha = mean)) +
    geom_errorbar(width = 0.2)