问题

I've created this function that puts together a list from object names, but can handle when the object doesn't exist. The function works fine on its own, but when I try and call it within another function I just get NA out for each list element. What's going on? Here's a simplified example:

我创建了这个函数，它可以从对象名称组合列表，但可以处理对象不存在的情况。这个函数单独运行正常，但当我尝试在另一个函数中调用它时，每个列表元素都变成了NA。发生了什么？以下是一个简化的示例：

genes <- c("gene1", "gene2", "gene3", "gene4")
analyses <- c("analysis_one", "analysis_two")

Function to put together a list even if an element doesn't exist

用于组合列表的函数，即使元素不存在也可以处理

append_if_exists <- function(element_names) {

Create empty list

创建一个空列表

my_list <- list()

Get and append each new element

获取并附加每个新元素

for (element_name in element_names) {

# Get new element
# 获取新元素
if (exists(element_name)) {
  element <- list(get(element_name))
  names(element) <- element_name
} else {
  element <- list(NA)
  names(element) <- element_name
}
# Append new element
# 附加新元素
my_list <- c(my_list, element)

}

Output

输出

return(my_list)

}

my_function <- function(genes, analyses) {

Analysis one

分析一

if ("analysis_one" %in% analyses) {
output1 <- genes[1:3]
}

Analysis two

分析二

if ("analysis_two" %in% analyses) {
output2 <- genes[1:2]
}

Append if exists

如果存在则附加

out <- append_if_exists(c("output1", "output2"))
return(out)

}

Run the function and my list has the right names, but the elements are just 'NA'

运行函数，我的列表有正确的名称，但元素只是'NA'

result <- my_function(genes, analyses)

Run the same thing outside of the function and it works fine!

在函数外运行相同的内容，它正常工作！

output1 <- genes[1:3]
output2 <- genes[1:2]
result <- append_if_exists(c("output1", "output2"))

genes <- c("gene1", "gene2", "gene3", "gene4")
analyses <- c("analysis_one", "analysis_two")
# Function to put together a list even if an element doesn't exist
append_if_exists <- function(element_names) {
  
  # Create empty list
  my_list <- list()
  
  # Get and append each new element
  for (element_name in element_names) {
    
    # Get new element
    if (exists(element_name)) {
      element <- list(get(element_name))
      names(element) <- element_name
    } else {
      element <- list(NA)
      names(element) <- element_name
    }
    
    # Append new element
    my_list <- c(my_list, element)
    
  }
  
  # Output
  return(my_list)
  
}
my_function <- function(genes, analyses) {
  
  # Analysis one
  if ("analysis_one" %in% analyses) {
    output1 <- genes[1:3]
  }
  
  # Analysis two
  if ("analysis_two" %in% analyses) {
    output2 <- genes[1:2]
  }
  
  # Append if exists
  out <- append_if_exists(c("output1", "output2"))
  return(out)
  
}
# Run the function and my list has the right names, but the elements are just 'NA'
result <- my_function(genes, analyses)
# Run the same thing outside of the function and it works fine!
output1 <- genes[1:3]
output2 <- genes[1:2]
result <- append_if_exists(c("output1", "output2"))

答案1

得分: 1

不需要回答我要翻译的问题。以下是翻译的内容：

与编写自己的append_if_exists不同，只需使用内置的mget（）函数。如果要避免不存在的名称，可以将名称与当前范围中的名称相交。这应该可以工作。

问题在于您的解决方案需要注意对象被找到的环境。R使用词法作用域，因此函数在定义函数的环境中解析变量，而不是在调用函数的环境中。如果您在全局环境中定义了append_if_exists，那么它无法看到在my_function运行时环境中定义的output1和output2。get（）函数确实有一个envir=参数，所以您可以将适当的环境传递给该函数，但这可能会很快变得混乱。

my_function <- function(genes, analyses) {
  ...
  out <- mget(intersect(c("output1", "output2"), ls()))
  return(out)
}