Question

是否有办法在R中使用dplyr打印分组平均值和sds（即在许多随机对照试验报告中观察到的情况，其中报告了安慰剂和药物组的基线人口统计数据）。目前我有：

group_summary<- df %>% 
  group_by(group) %>% 
  summarise_at( vars(age, iq), c(mean, sd), na.rm=TRUE)

t(group_summary)

将sds置于手段之下：

                                [,1]       [,2]       [,3]        
group                           "drug1"     "drug2"       "placebo"
age (x, ...) ...             "22.72436" "25.06164" "23.54394"  
weight (x, ...) ...           "57.97143" NA         "64.66667"  
age (x, na.rm = FALSE) ...   "3.793276" "3.930634" "4.623405"  
weight (x, na.rm = FALSE) ... "16.03617" NA         "15.44099"

理想情况下，我想要的更像是：

                            [,1]       [,2]       [,3]        
group                           "drug1"     "drug2"       "placebo"
age (x, ...) ...             "22.7(3.8)" "25.1(3.9)" "23.5(4.6)"  
weight (x, ...) ...           "56.0(16.0)" NA         "64.7(15.4)"

感谢任何建议，尽管如果可能的话，整理方法将是理想的

Answer 1

我用这个：

df1 <- df %>%
    group_by(grouping.variable) %>%
    summarise(#x = variable, y = number of decimal places
    mean.sd.x = paste0(format(round(mean(x), y), 
        nsmall = y),"(",format(round(sd(x), y), nsmall = y) ,")"))

Answer 2

你必须建立@ Pete的答案。如果你想为a，b，c和d这几个变量做这个，那就做吧..

and

我不太喜欢library(tidyverse) df%>% group_by(grouping.variable)%>% summarise_at(vars(a,b,c,d), funs(paste0(format(round(mean(., na.rm = T), y), nsmall = y),"(",format(round(sd(., na.rm = T), y), nsmall = y) ,")")))%>% gather(Variable, Value, -grouping.variable)%>% spread(grouping.variable, Value)的工作原理。我想使用t()包中的gather和spread。

我希望这会给你你想要的东西。

使用dply

2 个答案: