如何从t-test和柱状图的数据列表中选择特定的两组？

Question

如何从数据中选择两组进行t检验和条形图？我有几个组/衬底(即O1、O2、O3、S_O4、S_O5等)组成的数据，我需要使用特定的两个组/衬底(例如O1和S_O5)绘制条形图，然后运行t测试。我需要帮助。请检查附件中Excel数据的链接，好吗？

library(ggpubr)
library(rstatix)
library(xlsx)
library(patchwork)
df<-read.xlsx("umar_02.xlsx", header = T, 1)
# Statistical test

######## Figure 1 ########
stat.test <- df %>%
  t_test(Moisture_content ~ substrate) %>%
  add_significance()
stat.test

# Box plots with p-values
bxp1 <- ggboxplot(df, x = "substrate", y = "Moisture_content", fill = "substrate", 
                 palette = c("#00AFBB", "#E7B800"), width = 0.35)



stat.test <- stat.test %>% add_xy_position(x = "substrate")
bxp1 + 
  stat_pvalue_manual(stat.test, label = "p = {p}") +
  scale_y_continuous(expand = expansion(mult = c(0.05, 0.1)))+
  theme(
    aspect.ratio = 3
  )

Excel数据：enter link description here

Answer 1

df <- rio::import("https://docs.google.com/spreadsheets/d/1YZQRihyc5aTWehda8Wr24xwm23a87jv7/edit#gid=1007932109")
library(ggpubr)
library(rstatix)
stat.test <- df %>%
  filter(Ortamlar %in% c("O1", "S_O5")) %>% 
  t_test(Moisture_content ~ Ortamlar) %>%
  add_significance()
stat.test

stat.test
# # A tibble: 1 x 9
#   .y.              group1 group2    n1    n2 statistic    df     p p.signif
# * <chr>            <chr>  <chr>  <int> <int>     <dbl> <dbl> <dbl> <chr>   
# 1 Moisture_content O1     S_O5       3     3     -2.20  2.50 0.132 ns      

# Box plots with p-values
bxp1 <- df %>%
  filter(Ortamlar %in% c("O1", "S_O5")) %>%
  ggboxplot( x = "Ortamlar", y = "Moisture_content", fill = "Ortamlar", 
                  palette = c("#00AFBB", "#E7B800"), width = 0.35, xlab = "Substrate")


stat.test %>% add_xy_position(x = "Ortamlar")
stat.test <- stat.test %>% mutate(y.position=5.15)
bxp1 + 
  stat_pvalue_manual(stat.test, label = "p = {p}") +
  scale_y_continuous(expand = expansion(mult = c(0.05, 0.1)))+
  theme(
    aspect.ratio = 3
  )

​

​

Answer 2

我们可能需要使用%in%进行filter

library(dplyr)
stat.test <- df %>% 
  filter(Ortamlar %in% c("O1", "S_O5")) %>%
  rename(substrate = Ortamlar) %>%
  t_test(Moisture_content ~ substrate) %>%
  add_significance()

-output

stat.test
# A tibble: 1 x 9
  .y.              group1 group2    n1    n2 statistic    df     p p.signif
* <chr>            <chr>  <chr>  <int> <int>     <dbl> <dbl> <dbl> <chr>   
1 Moisture_content O1     S_O5       3     3     -2.20  2.50 0.132 ns