如何从多层嵌套列表提取信息并转换为DataFrame？

Got it, let's solve this nested list to DataFrame problem—this is such a common pain point with hierarchical data, especially when dealing with family trees or nested API outputs! The key here is to use a recursive function to traverse every level of the nested list, extract each person's attributes, and combine them all into a flat table.

Step 1: Build a Recursive Function with tidyverse #

We'll create a function that handles both single family members and lists of members, recursively digging into any children sublists:

1. Extracts all attributes except children for the current member
2. Recursively processes any nested children
3. Combines the current member's data with their children's data into a single table

Here's the code:

library(tidyverse)

flatten_family <- function(member_list) {
  # If input is a list of members, map the function over each one
  if (is.null(names(member_list))) {
    return(map_dfr(member_list, flatten_family))
  }
  
  # Pull out the current member's attributes (exclude children)
  current_member <- member_list %>%
    keep(~!is.list(.x) || names(.x) != "children") %>%
    as_tibble_row()
  
  # Process children if they exist
  if (!is.null(member_list$children)) {
    children_data <- map_dfr(member_list$children, flatten_family)
    bind_rows(current_member, children_data)
  } else {
    # No children? Just return the current member's data
    current_member
  }
}

Step 2: Apply the Function to Your family List #

Run the function on your nested list, and you'll get a flat DataFrame where each row represents one family member. Missing attributes (like hair for most people) will automatically be filled with NA:

family_df <- flatten_family(family)

# View the full result
print(family_df, n = Inf)

What This Does #

• Handles Any Nesting Depth: The function calls itself on every children sublist, so it works even if Harry grows up and has kids later!
• Flexible Attributes: as_tibble_row and bind_rows handle members with different attribute sets seamlessly—no errors if someone has a hair field and others don't.
• Clean, Flat Output: Every person in the family tree gets their own row with all their personal details.

Example Output Snippet #

You'll see a table like this (truncated for brevity):

# A tibble: 11 × 4
   name    age eyes  hair 
   <chr> <dbl> <chr> <chr>
 1 Alice    40 blue  NA   
 2 Bob      20 blue  NA   
 3 Charlie  18 brown NA   
 4 Dan      12 green NA   
 5 Erin     69 green NA   
 6 Frank    45 blue  NA   
 7 George   24 blue  NA   
 8 Harry     2 green NA   
 9 Ingrid   22 brown brown
10 Jack     29 brown NA   
11 Larry    21 blue  NA   

Alternative for Shallow Nesting (Not Recommended for Arbitrary Depth) #

If you knew the nesting level was fixed, you could use repeated unnest calls, but this gets messy fast with deeper trees:

tibble(family = family) %>%
  unnest_wider(family) %>%
  unnest_longer(children, keep_empty = TRUE) %>%
  unnest_wider(children, names_sep = "_")

Stick with the recursive function for true flexibility with any nesting depth.

内容的提问来源于stack exchange，提问作者CDoug