在R语言中基于多条件为数据框分配新列值的实现方法

Got it, let's tackle this problem. You need to add a new column to your data frame based on multiple custom rules, and the solution should be flexible enough to adapt as you add more rules later. Here's a straightforward, scalable approach using base R or dplyr:

First, let's start with some sample data matching your a1-a9 column structure:

set.seed(123) # For reproducible results
df <- data.frame(
  a1 = sample(1:9, 10, replace = TRUE),
  a2 = sample(1:9, 10, replace = TRUE),
  a3 = sample(1:9, 10, replace = TRUE),
  a4 = sample(1:9, 10, replace = TRUE),
  a5 = sample(1:9, 10, replace = TRUE),
  a6 = sample(1:9, 10, replace = TRUE),
  a7 = sample(1:9, 10, replace = TRUE),
  a8 = sample(1:9, 10, replace = TRUE),
  a9 = sample(1:9, 10, replace = TRUE)
)

Step 1: Define Your Rules as a List #

The key to scalability is storing rules in a structured list. Each rule includes a condition (what triggers it) and the value to assign to the new column when the condition is met. We'll also add a default value for rows that don't match any rule.

For your example rules, structure them like this (adjust the order to set rule priority—earlier rules get checked first):

rules <- list(
  # Rule 1: If any column has 1 or 2, assign 1
  rule1 = list(
    condition = c(1, 2),
    value = 1
  ),
  # Rule 2: If any column has 5 or 6, assign 6 (only checked if Rule 1 fails)
  rule2 = list(
    condition = c(5, 6),
    value = 6
  ),
  # Default value if no rules are matched
  default = list(
    value = 3
  )
)

Step 2: Create a Rule Application Function #

Write a helper function that checks each row against your rules in order and returns the appropriate value:

apply_rules <- function(row, rules) {
  # Loop through all rules except the default
  for (rule in setNames(rules, NULL)[-length(rules)]) {
    # Check if any value in the row matches the rule's condition
    if (any(row %in% rule$condition)) {
      return(rule$value)
    }
  }
  # If no rules are triggered, return the default value
  return(rules$default$value)
}

Step 3: Apply the Rules to Your Data Frame #

You can use either base R or dplyr to apply this function to every row:

Option 1: Base R #

df$NEW <- apply(df, 1, apply_rules, rules = rules)

Option 2: dplyr (tidyverse style) #

If you prefer the tidyverse workflow:

library(dplyr)

df <- df %>%
  rowwise() %>%
  # Check all columns from a1 to a9
  mutate(NEW = apply_rules(c_across(a1:a9), rules)) %>%
  ungroup()

Step 4: View the Result #

Print your updated data frame to see the new NEW column:

print(df)

Sample output:

a1 a2 a3 a4 a5 a6 a7 a8 a9 NEW
1   3  2  7  6  9  6  5  8  1   1
2   8  7  8  9  7  7  9  9  2   1
3   4  9  5  5  3  3  7  3  5   6
4   8  1  9  8  1  1  3  1  9   1
5   9  5  2  6  5  2  8  2  7   1
6   1  3  6  2  8  9  1  4  4   1
7   6  8  3  3  6  8  2  5  8   1
8   5  6  1  1  2  4  6  7  3   1
9   8  4  4  4  4  5  4  6  6   6
10  9  1  7  7  7  6  9  9  8   1

Scaling to More Complex Rules #

This approach is super flexible. If you need more complex conditions (like checking if the mean of the row is above a threshold, or all columns meet a criteria), update your rules to use functions instead of value lists:

# Example with a mix of condition types
rules <- list(
  rule_high_avg = list(
    condition = function(row) mean(row) > 6, # Custom function condition
    value = "high_average"
  ),
  rule_has_low_values = list(
    condition = c(1, 2), # Value match condition
    value = "has_1_or_2"
  ),
  default = list(
    value = "other"
  )
)

# Updated function to handle both condition types
apply_rules <- function(row, rules) {
  for (rule in setNames(rules, NULL)[-length(rules)]) {
    if (is.function(rule$condition)) {
      # Evaluate the custom function
      if (rule$condition(row)) {
        return(rule$value)
      }
    } else {
      # Check for value matches
      if (any(row %in% rule$condition)) {
        return(rule$value)
      }
    }
  }
  return(rules$default$value)
}

Now you can add, remove, or reorder rules in the rules list without changing the core application code—perfect for managing multiple custom conditions!

内容的提问来源于stack exchange，提问作者C_psy