R语言中拆分含多值列失败问题求助

Hey there! Let's figure out how to split that exercise barriers column in your dataset—since the entries use | (with surrounding spaces) as a delimiter and have varying numbers of values, here are two practical approaches depending on what you need:

1. Split into columns based on value position #

If you want each value in its own column ordered by where it appears in the original string (e.g., first value in column 1, second in column 2, etc.), use tidyr::separate_wider_delim—it handles varying lengths automatically:

First, let's recreate your sample data to test with:

# Sample dataset matching your example
exercise_data <- data.frame(
  barriers = c(
    "No transportation",
    "No transportation | Physically Unable | Don't know how",
    "Physically Unable | No transportation",
    "No Money | Don't know how | No Facilities | No free time"
  )
)

Now run the split:

library(tidyr)

# Split the barriers column into multiple position-based columns
split_by_position <- exercise_data %>%
  separate_wider_delim(
    cols = barriers,
    delim = " | ",  # Important: matches the exact delimiter (space + | + space)
    names_sep = "_",  # Names new columns like barriers_1, barriers_2, etc.
    too_few = "align_start",  # Shorter entries fill the first columns, leave others blank
    too_many = "error"  # Throws error if any entry has more values than the max found
  )

This will create columns barriers_1 through barriers_4 (since your longest entry has 4 values), with each column holding the corresponding reason from the original string.

2. Split into one-hot encoded columns (each unique reason as a column) #

If you want a separate column for every unique barrier (marked 1 if present, 0 if not), combine separate_longer_delim and pivot_wider:

library(tidyr)
library(dplyr)

# Convert barriers to one-hot encoded columns
one_hot_barriers <- exercise_data %>%
  # First, split the string into individual rows per barrier
  separate_longer_delim(cols = barriers, delim = " | ") %>%
  # Add a marker that the barrier is present
  mutate(is_present = 1) %>%
  # Pivot back to wide format, filling missing values with 0
  pivot_wider(
    names_from = barriers,
    values_from = is_present,
    values_fill = 0
  )

This will give you columns like No transportation, Physically Unable, Don't know how, etc., where each row has a 1 if that barrier was listed, and 0 otherwise.

Just a quick note: double-check your delimiter—if some entries don't have spaces around the |, adjust the delim argument to just "|" and add a str_trim() step (from the stringr package) to clean up leftover whitespace!

内容的提问来源于stack exchange，提问作者Christopher Williams