0

尽管使用了函数变体，但我在一个函数中使用多个 dplyr 函数时遇到问题。

例子

library(dplyr)

# Data:
mydf <- data.frame( 
  var1     = factor(rep(1:24, each = 100)),
  var2    = runif(2400, min = -10, max = 125),
  var3     = runif(2400, min = 0, max = 2500),
  var4     = runif(2400, min = - 10, max = 25)  
)

# The function I want to build: 


fx.average <- function(df, varlist) {
  # select some varibles from a data frame
  df <- df %>% dplyr::select_(.dots = varlist)

   # Group by a variable and then just calculate the mean
  df <-df %>% dplyr::group_by_(var1) %>% # added df here
    dplyr::summarise_each_(funs_(mean(., na.rm = TRUE)))
}

所以，现在我要测试以下功能：

# Test function, Setup var-list
varlist0 <- c("var1", "var2", "var3")    

fx.average(mydf, varlist0)
# Error in dplyr::group_by_(var1) : object 'var1' not found
#  object 'var1' not found 


# Manual example
mydf %>% dplyr::select(var1, var2, var3) %>% 
  group_by(var1) %>% 
  summarise_each(funs(mean(., na.rm = TRUE)))

不知道出了什么问题？从其他问题来看，似乎应该通过在函数中添加下划线来解决 - 因为它们是为在函数内部使用而构建的？

4

1 回答 1

1

在 OP 的代码中，有一些拼写错误（未指定group_by步骤中的数据以及使用不带引号的字符串的 NSE 以及使用funs_wheresummarise_each_和summarise_eachworks funs）

fx.average <- function(df, varlist) {
  df %>% 
      dplyr::select_(.dots = varlist) %>%
      dplyr::group_by_(.dots = "var1") %>% 
      dplyr::summarise_each(funs(mean(., na.rm = TRUE)))
  }

fx.average(mydf, varlist0)
# A tibble: 24 × 3
#     var1     var2     var3
#   <fctr>    <dbl>    <dbl>
#1       1 55.13601 1141.021
#2       2 59.16508 1155.226
#3       3 59.64524 1245.043
#4       4 60.12310 1284.808
#5       5 57.65874 1221.771
#6       6 58.86611 1266.026
#7       7 66.13987 1303.927
#8       8 54.21595 1303.638
#9       9 63.84230 1280.380
#10     10 49.15238 1236.456
# ... with 14 more rows
于 2017-02-20T14:38:18.920 回答