R语言函数开发：生成指定排序的命名数值列表（基于查找表）

Hey there! Let's build this R function step by step to meet all your requirements. First, let's clarify the core needs, then dive into the code and adjustments.

• The output q.arg2 must match the length of the input character vector factors2
• For each name in factors2:
  • If its distribution type in dist2 is "normal", extract the corresponding mean and sd values from dist2 and return a sub-list like list(mean=x, sd=y)
  • If it's another distribution (e.g., "weibull"), calculate the shape and scale parameters using the corresponding observations in data frame k2, then return list(shape=x, scale=y)
• Finally, sort q.arg2 according to the order specified in sort2

First, let's fix the dist2 data frame from your reference code—its original row-based structure isn't easy to query. We'll restructure it so each variable (var1/var2/var3) is a row with its corresponding id, mean, sd, and distribution:

# Generate dist2 correctly
id <- c("a", "b", "c")
mean <- c(3, 1, 4)
sd <- c(2,7,8)
distribution <- c("weibull", "normal", "normal")
dist2 <- data.frame(id, mean, sd, distribution)
rownames(dist2) <- c("var1", "var2", "var3") # Map rows to variable names
dist2$mean <- as.numeric(dist2$mean)
dist2$sd <- as.numeric(dist2$sd)

# Generate factors2
factors2 <- c("var2", "var3")

# Generate k2 (data for weibull fitting, fixed rnorm parameter order)
set.seed(123) # For reproducible results
k2 <- data.frame(replicate(2, rnorm(n=300, mean=100, sd=2)))
colnames(k2) <- c("var1", "var2")

# Load required package
library(fitdistrplus)

# Sort order (matches your reference "a,b" which maps to var1/var2)
sort2 <- "a,b"

This function handles both variable-name and id-based sorting, and includes error handling for missing variables or unsupported distributions:

generate_qarg2 <- function(factors2, dist2, k2, sort2) {
  # Handle sort2: if it uses ids (like "a,b"), map to variable names
  sort_ids <- strsplit(sort2, ",")[[1]]
  sort_order <- rownames(dist2)[match(sort_ids, dist2$id)]
  
  # Initialize empty list to store results
  q.arg2 <- list()
  
  # Iterate over each variable in factors2
  for (var in factors2) {
    # Get distribution info for the current variable
    dist_info <- dist2[var, ]
    
    if (dist_info$distribution == "normal") {
      # Extract pre-defined mean and sd (ensured numeric)
      q.arg2[[var]] <- list(mean = dist_info$mean, sd = dist_info$sd)
    } else if (dist_info$distribution == "weibull") {
      # Check if variable exists in k2 before fitting
      if (!var %in% colnames(k2)) {
        stop(paste("Variable", var, "not found in k2 data frame"))
      }
      # Fit weibull distribution and extract parameters
      fw <- fitdist(k2[[var]], "weibull")
      q.arg2[[var]] <- list(shape = fw$estimate["shape"], scale = fw$estimate["scale"])
    } else {
      # Handle unsupported distribution types
      stop(paste("Unsupported distribution type:", dist_info$distribution))
    }
  }
  
  # Sort the list: keep variables in sort_order first, then add remaining ones
  sorted_vars <- intersect(sort_order, factors2)
  sorted_vars <- c(sorted_vars, setdiff(factors2, sorted_vars))
  q.arg2 <- q.arg2[sorted_vars]
  
  # Remove top-level names to match your reference q.arg2 format
  names(q.arg2) <- NULL
  
  return(q.arg2)
}

# Run the function with your sample data
q.arg2_result <- generate_qarg2(factors2, dist2, k2, sort2)

# View the output
print(q.arg2_result)

Sample output (matches your reference format):

[[1]]
[[1]]$mean
[1] 1

[[1]]$sd
[1] 7


[[2]]
[[2]]$mean
[1] 4

[[2]]$sd
[1] 8

• Data Structure Fix: We restructured dist2 to make variable-specific queries straightforward, and converted mean/sd to numeric types (they were stored as characters in your original code).
• Distribution Handling: For normal distributions, we pull pre-defined parameters directly from dist2. For weibull distributions, we use fitdist() from fitdistrplus to estimate shape/scale from k2.
• Sorting Logic: The function maps sort2 ids to variable names (matching your reference "a,b" input), then reorders the result list to match this order. Any variables in factors2 not in sort2 are added to the end.
• Error Handling: Includes checks for missing variables in k2 and unsupported distribution types to avoid unexpected crashes.

内容的提问来源于stack exchange，提问作者Avery