“填充”内箱线图中的重要性标签

Question

我正在为一次会议绘制一个图，并想添加更多的意义栏。这是我的代码：

bxp_stress <-
  subjstress %>%
  ggboxplot( x = "timetask",
             y = "Stress",
             fill = "group",
             xlab = "timepoint",
             ylab = "subjective rating [VAS]",
             title = "Subjective stress"
  )+
  scale_fill_manual(values = my_palette) +
  scale_x_discrete(labels = c("pre-ctrl", "post-ctrl", "pre-stress", "post-stress")) +
  geom_signif(comparisons = list(c("MAST_pre", "MAST_post"), c("plac_post", "MAST_post")),
                                 annotations = "***",
              step_increase = 0.05,
              map_signif_level = TRUE,
              vjust = 0.4, 
              tip_length = 0)

我的情节看起来如何：

现在我想为“三组之间的后应力”添加显着性条（在 ggboxplot 的“fill”参数中定义）。 我该如何做到这一点，因为仅在 2 个组（NC 与 IUD）之间发现显着性，并且我不想要整体显着性，因为我正在使用以下代码：

bxp_stress + stat_compare_means(aes(group = group), label = "p.signif")

编辑：这就是我的想象：

编辑2：可用的数据框

subjstress <- tibble(
  group = rep(c("NC", "IUD", "OC"), each = 232),
  timetask = rep(c("plac_pre", "plac_post", "MAST_pre", "MAST_post"), times = 174),
  Stress = sample(0:100, 696, replace = TRUE)
)

Answer 1

您可以首先使用

add_significance

和 t 检验创建数据框来绘制显着性。之后，您可以通过使用

filter

修改绘制的图层，仅绘制所需的组，如下所示：ggplot_build

library(ggpubr)
library(dplyr)
library(rstatix)

stat.test <- subjstress %>%
  group_by(timetask) %>%
  t_test(Stress ~ group) %>%
  adjust_pvalue() %>%
  add_significance("p.adj") %>%
  add_xy_position(x = "timetask") 
stat.test
#> # A tibble: 12 × 16
#>    timetask  .y.    group1 group2    n1    n2 statistic    df     p p.adj
#>    <chr>     <chr>  <chr>  <chr>  <int> <int>     <dbl> <dbl> <dbl> <dbl>
#>  1 MAST_post Stress IUD    NC        58    58    0.998   113. 0.32      1
#>  2 MAST_post Stress IUD    OC        58    58    0.840   113. 0.402     1
#>  3 MAST_post Stress NC     OC        58    58   -0.159   114. 0.874     1
#>  4 MAST_pre  Stress IUD    NC        58    58    0.851   114. 0.396     1
#>  5 MAST_pre  Stress IUD    OC        58    58    1.51    113. 0.134     1
#>  6 MAST_pre  Stress NC     OC        58    58    0.649   114. 0.518     1
#>  7 plac_post Stress IUD    NC        58    58   -0.142   113. 0.888     1
#>  8 plac_post Stress IUD    OC        58    58   -0.581   114. 0.562     1
#>  9 plac_post Stress NC     OC        58    58   -0.415   114. 0.679     1
#> 10 plac_pre  Stress IUD    NC        58    58   -1.09    114. 0.279     1
#> 11 plac_pre  Stress IUD    OC        58    58    0.0732  114. 0.942     1
#> 12 plac_pre  Stress NC     OC        58    58    1.17    114. 0.246     1
#> # ℹ 6 more variables: p.adj.signif <chr>, y.position <dbl>,
#> #   groups <named list>, x <dbl>, xmin <dbl>, xmax <dbl>

p <- subjstress %>%
  ggboxplot( x = "timetask",
             y = "Stress",
             fill = "group",
             xlab = "timepoint",
             ylab = "subjective rating [VAS]",
             title = "Subjective stress"
  ) +
  stat_pvalue_manual(
    stat.test,  label = "p.adj.signif", tip.length = 0.01
  ) + 
  scale_x_discrete(labels = c("pre-ctrl", "post-ctrl", "pre-stress", "post-stress")) 
p

创建于 2023-08-18，使用

reprex v2.0.2