如何将两组配对t检验结果转为Word表格？R语言报错求助

问题说明 #

我尝试将两次配对t检验的结果整合为单个表格并导出到Microsoft Word，但触发错误：Error: (converted from warning) package ‘flextable’ was built under R version 4.2.3。虽然已经更新了R版本，但还是出现这个版本兼容问题。

现有代码 #

原始配对t检验代码 #

## 检验边缘存在对鸟类捕食标记量的影响
library(tidyverse)
# 筛选样地位置为Edge和Center的数据
filtered_data <- subset(Yes, `Position within the plot` %in% c("Edge", "Center"))

print(result <- t.test(`Amount of bird markings per caterpillar` ~ `Position within the plot`, data = filtered_data, paired = TRUE))

## 检验边缘缺失对鸟类捕食标记量的影响
# 筛选样地位置为Edge和Center的数据
filtered_datan <- subset(no, `Position within the plot` %in% c("Edge", "Center"))

print(resultn <- t.test(`Amount of bird markings per caterpillar` ~ `Position within the plot`, data = filtered_datan, paired = TRUE))

尝试过的导出代码 #

result <- data.frame(
  statistic = result$statistic,
  df = result$parameter,
  p_value = result$p.value,
  conf_int_lower = result$conf.int[1],
  conf_int_upper = result$conf.int[2],
  mean_difference = result$estimate
)

# 导出为CSV
write.csv(result, file = "t_test_results.csv", row.names = FALSE)
tab_model(model, file = "t_test_results.doc", show.se = TRUE, show.std = TRUE, show.stat = TRUE, show.ci = 0.95, show.est = TRUE)

library(flextable)

# 读取CSV并转换为flextable
result <- read.csv("t_test_results.csv")
ft <- flextable(result)

# 修改列名
colnames(ft) <- c("Statistic", "df", "p-value", "Conf. Int. Lower", "Conf. Int. Upper", "Mean Difference")

解决步骤 #

1. 修复flextable版本兼容问题 #

更新R后，旧版本的flextable是基于旧R版本编译的，必须重新安装适配当前R版本的包：

install.packages("flextable")

如果提示需要从源码安装，选择同意，这样包会重新编译适配你的当前R版本。

2. 优化结果整合与导出流程 #

用broom包简化结果提取，合并两次检验结果后直接用flextable导出到Word：

library(tidyverse)
library(broom)
library(flextable)

# 处理第一次检验（有边缘组）
filtered_data <- subset(Yes, `Position within the plot` %in% c("Edge", "Center"))
result <- t.test(`Amount of bird markings per caterpillar` ~ `Position within the plot`, data = filtered_data, paired = TRUE)
result_tidy <- tidy(result) %>% mutate(检验组 = "有边缘组")

# 处理第二次检验（无边缘组）
filtered_datan <- subset(no, `Position within the plot` %in% c("Edge", "Center"))
resultn <- t.test(`Amount of bird markings per caterpillar` ~ `Position within the plot`, data = filtered_datan, paired = TRUE)
resultn_tidy <- tidy(resultn) %>% mutate(检验组 = "无边缘组")

# 合并两次结果并整理列名
combined_results <- bind_rows(result_tidy, resultn_tidy) %>%
  select(检验组, statistic, df, p.value, conf.low, conf.high, estimate) %>%
  rename(
    "统计量(t值)" = statistic,
    "自由度(df)" = df,
    "p值" = p.value,
    "置信区间下限" = conf.low,
    "置信区间上限" = conf.high,
    "均值差" = estimate
  )

# 生成表格并导出到Word
ft <- flextable(combined_results) %>%
  set_caption("配对t检验结果：边缘存在与否对鸟类捕食标记量的影响") %>%
  autofit()

save_as_docx(ft, path = "t检验结果汇总.docx")

注意事项 #

• 确保Yes和no两个数据框存在，且列名与代码中的完全匹配；
• 如果Windows系统安装flextable时编译失败，需要先安装Rtools；Mac系统则需要安装Xcode命令行工具。

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