如何在R语言中使用While循环对对应列进行相关性分析

Hey there! Great question—using ncol() to anchor your loop logic is exactly the right approach here. Let's walk through how to set this up properly, with steps that are easy to follow and adapt to your data.

First, let's make some concrete assumptions to keep things straightforward:

• You've already imported your Excel files into R as data frames, let's call them df_1997 (for file A) and df_2000 (for file B).
• Both data frames have exactly 35 columns (as you noted), with matching column positions (i.e., column 1 in both refers to the same dietary variable, column 2 the same, etc.).

Step 1: Prep Your Loop Setup #

First, we'll initialize our loop counter, grab the total number of columns, and create a storage object to keep track of all our correlation results (this is way better than printing results to the console one by one):

# Initialize loop counter
i <- 1

# Get total number of columns (use either data frame since they're identical in column count)
total_cols <- ncol(df_1997)

# Create an empty list to store correlation test results
cor_results <- list()

# Optional: Add a safety check to ensure column counts match
if (ncol(df_1997) != ncol(df_2000)) {
  stop("Error: The two data frames have different numbers of columns!")
}

Step 2: Build the While Loop #

Now we'll write the loop that iterates through each column pair, runs the correlation test, and stores the result:

while (i <= total_cols) {
  # Extract the current column from each data frame
  current_col_1997 <- df_1997[, i]
  current_col_2000 <- df_2000[, i]
  
  # Run a correlation test (Pearson by default; adjust method = "spearman" or "kendall" if needed)
  test_result <- cor.test(current_col_1997, current_col_2000)
  
  # Store the result in our list, using the column name as the list item name for clarity
  names(cor_results)[i] <- colnames(df_1997)[i]
  cor_results[[i]] <- test_result
  
  # Increment the counter to avoid an infinite loop
  i <- i + 1
}

Step 3: Explore Your Results #

Once the loop finishes, you can easily access and organize the results:

• To view the full test result for a specific column (e.g., column "A"):

  cor_results[["A"]]
  

• To create a clean summary table with key metrics (correlation coefficient, p-value, confidence interval):

  summary_table <- lapply(cor_results, function(x) {
    data.frame(
      Correlation_Coefficient = x$estimate,
      P_Value = x$p.value,
      CI_Lower = x$conf.int[1],
      CI_Upper = x$conf.int[2]
    )
  })
  
  # Convert the list to a single, easy-to-read data frame
  do.call(rbind, summary_table)
  

A Quick Side Note on Alternatives #

While a while loop works perfectly here, you might find a for loop more intuitive for this fixed-column iteration (the logic is nearly identical). If you ever want to switch, just replace the while block with:

for (i in 1:total_cols) {
  # Same code as inside the while loop here
}

内容的提问来源于stack exchange，提问作者MT32