How to show the result of compare_means() in the stat_compare_means()

[LiZhihua1982]

Hi Everyone,
My data is below

compare_means(value~group2, group.by="variable",data=fm,method = "kruskal.test")

A tibble: 12 × 7

variable .y. p p.adj p.format p.signif method

1 Tyramine value 0.000349 0.0042 0.00035 *** Kruskal-Wallis
2 Cadaverine value 0.000349 0.0042 0.00035 *** Kruskal-Wallis
3 Histamine value 0.0243 0.097 0.02434 * Kruskal-Wallis
4 Putrescine value 0.508 0.8 0.50780 ns Kruskal-Wallis
5 Ornithine value 0.000349 0.0042 0.00035 *** Kruskal-Wallis
6 Tryptamine value 0.0193 0.096 0.01928 * Kruskal-Wallis
7 Spermidine value 0.0380 0.11 0.03798 * Kruskal-Wallis
8 Phenethylamine value 0.00232 0.016 0.00232 ** Kruskal-Wallis
9 X3.Methoxytyramine value 0.00232 0.016 0.00232 ** Kruskal-Wallis
10 Ethanolamine value 0.000675 0.0061 0.00067 *** Kruskal-Wallis
11 Spermine value 0.402 0.8 0.40154 ns Kruskal-Wallis
12 Agmatine value 0.00127 0.01 0.00127 ** Kruskal-Wallis
.............

when I use this order :p <- ggline(fm, x = "time", y = "value", color = "group2",
shape = 21,point.size =2,stroke =1,
add = c("mean_sd", "jitter"),
xlab = 'Days',ylab = 'value',
palette = c("#f94040", "#0000ff"),
facet.by=c("variable"))+theme_set(theme_bw())

when I use this order
p2<- p+stat_compare_means(aes(group = group2),method = "kruskal.test",label = "p.signif",size=2.8,label.x = 1.4,label.y = 8000000)

When I use this order: p+stat_compare_means(), it shows the comparsion with different "time" not the "group2"

So, As shown in compare_means(), Tyramine is significant which I want to show it in the stat_compare_means(),I don't know how to change the order, would you help me? Thanks