如何批量处理前缀为points_的考试得分变量转换为0/1对错标识？

Got it, let's put an end to that tedious repetitive coding for your exam score variables! Since all your target variables follow the points_* naming pattern, we can use batch variable handling in whatever tool you’re using—no more writing code for each points_18616-style variable one by one. Here are tailored solutions for the most common data analysis tools:

R (Using Tidyverse/dplyr) #

If you’re working in R with the tidyverse, dplyr’s across() function lets you target all points_ prefixed variables in one go:

library(dplyr)

# Replace `df` with your actual dataset name
df <- df %>%
  # Convert all points_* variables to 1 (if score > 0) or 0 (if score = 0)
  mutate(
    across(
      starts_with("points_"), 
      ~ifelse(.x > 0, 1, 0),
      # Optional: Rename new variables to "correct_*" instead of overwriting originals
      .names = "correct_{col}"
    )
  )

• If you want to overwrite the original points_* variables instead of creating new ones, just remove the .names argument.
• Adjust the condition (>.x > 0) if your "correct" definition differs (e.g., .x == full_score if only perfect answers count).

Stata #

In Stata, you can use a local macro to grab all matching variables, then loop through them:

// Step 1: Get a list of all variables starting with "points_"
local point_vars: list varlist & points_*

// Step 2: Loop through each variable to create 0/1 correct indicators
foreach var of local point_vars {
    // Create a new variable (correct_points_*) with 1 for non-zero scores, 0 otherwise
    gen correct_`var' = (`var' > 0)
    
    // Uncomment below to overwrite the original `points_*` variable instead:
    // replace `var' = (`var' > 0)
}

Stata automatically converts logical expressions (var > 0) to 1/0 values, so this is concise and efficient.

Python (Using Pandas) #

For pandas users, filter the target columns and apply a vectorized transformation:

import pandas as pd

# Replace `df` with your dataset name
# Get all columns starting with "points_"
point_columns = df.filter(like="points_").columns

# Option 1: Create new "correct_*" columns
df[[f"correct_{col}" for col in point_columns]] = df[point_columns].applymap(lambda x: 1 if x > 0 else 0)

# Option 2: Overwrite the original `points_*` columns
# df[point_columns] = df[point_columns].applymap(lambda x: 1 if x > 0 else 0)

applymap() applies the lambda function to every element in the selected columns, making this a fast, batch operation.

All these approaches work across different datasets as long as the points_* naming rule holds—no need to tweak code for each new dataset!

内容的提问来源于stack exchange，提问作者DrL