如何在R语言中生成多选交叉表并去除行首的TRUE/FALSE标识

嘿，我之前用expss做交叉表的时候也遇到过这个烦人的问题！你现在的代码里，tab_cells()直接用了逻辑判断（比如qpd4_1==1），expss会默认把这些逻辑向量的两个取值（TRUE/FALSE）都作为行标签的一部分显示出来。要只保留你自定义的文本（比如"By phone"），可以试试下面两种方法：

方法1：先创建带标签的指标变量（更清晰易维护） #

这种方法先把每个逻辑判断转换成带自定义标签的变量，后续生成表格时就只会显示你设置的标签：

library(expss)
library(haven)

X4707 <- read_sav("/home/cfmc/4707/data/4707.sav")

# 创建指标变量并设置自定义标签
X4707 <- X4707 %>%
    mutate(
        by_phone = qpd4_1 == 1,
        by_email = qpd4_2 == 1,
        utility_website = qpd4_3 == 1,
        roseville_notification = qpd4_4 == 1,
        social_media = qpd4_5 == 1,
        text = qpd4_6 == 1,
        not_sure = qpd4_8 == 1
    ) %>%
    apply_labels(
        by_phone = "By phone",
        by_email = "By email",
        utility_website = "Utility website",
        roseville_notification = "Roseville Electric notification",
        social_media = "Social media",
        text = "Text",
        not_sure = "Not sure"
    )

# 生成交叉表，只计算TRUE对应的百分比并隐藏空行
X4707 %>% 
    tab_cells(by_phone, by_email, utility_website, roseville_notification, social_media, text, not_sure) %>% 
    tab_cols(total(), qf5) %>% 
    tab_stat_cpct(only_values = TRUE) %>%  # 只保留TRUE的统计值
    tab_last_sig_cpct() %>% 
    tab_pivot(drop_empty_rows = TRUE)

方法2：直接在tab_cells里处理（更简洁） #

如果你不想额外创建变量，可以直接在tab_cells()里用as.category()给逻辑值指定唯一的标签，同时过滤掉FALSE对应的空行：

library(expss)
library(haven)

X4707 <- read_sav("/home/cfmc/4707/data/4707.sav")

X4707 %>% 
    tab_cells(
        "By phone" = as.category(qpd4_1 == 1, labels = c("By phone" = TRUE)),
        "By email" = as.category(qpd4_2 == 1, labels = c("By email" = TRUE)),
        "Utility website" = as.category(qpd4_3 == 1, labels = c("Utility website" = TRUE)),
        "Roseville Electric notification" = as.category(qpd4_4 == 1, labels = c("Roseville Electric notification" = TRUE)),
        "Social media" = as.category(qpd4_5 == 1, labels = c("Social media" = TRUE)),
        "Text" = as.category(qpd4_6 == 1, labels = c("Text" = TRUE)),
        "Not sure" = as.category(qpd4_8 == 1, labels = c("Not sure" = TRUE))
    ) %>% 
    tab_cols(total(), qf5) %>% 
    tab_stat_cpct() %>% 
    tab_last_sig_cpct() %>% 
    tab_pivot(drop_empty_rows = TRUE)  # 去掉FALSE对应的空行

这两种方法都能让你的交叉表行区域（stub）只显示你设置的自定义文本，再也看不到TRUE/FALSE啦！

内容的提问来源于stack exchange，提问作者larryb