R循环中如何将生成的多个DataFrame存储到全局环境？

Hey there! Let's sort out this problem you're facing with splitting your 58-column DataFrame into 29 smaller ones. You're totally right about the root cause—your assign() call is using the fixed string "dfi" every time, so each loop iteration just overwrites the same variable instead of creating a unique one for each column pair.

Here are two solid solutions for you:

1. Dynamically Generate Unique Variable Names #

If you really want separate variables (like df1, df2, ..., df29) in your global environment, you need to create a unique name for each iteration using paste0() to combine "df" with the loop index i. Here's the adjusted code:

for (i in 1:29) {
  # Extract the i-th pair of columns
  x <- LCMS[, ((i*2)-1):(i*2)]
  # Create a unique variable name (df1, df2, ...) and assign the data
  assign(paste0("df", i), x, envir = globalenv())
}

This way, each loop creates a new variable instead of overwriting the same dfi every time. You'll end up with df1 holding columns 1&2, df2 holding 3&4, and so on up to df29.

2. Use a List (Recommended!) #

While separate variables work, using a list to store all your smaller DataFrames is a much cleaner approach in R. Lists let you easily manage, iterate over, and manipulate all your subsets without cluttering your global environment with 29 separate variables. Here's how to do it:

# Initialize an empty list to hold your split DataFrames
df_list <- list()

# Loop through each column pair and add to the list
for (i in 1:29) {
  df_list[[i]] <- LCMS[, ((i*2)-1):(i*2)]
}

# Optional: Name the list elements for easier access
names(df_list) <- paste0("df", 1:29)

# Access individual subsets like this:
df_list$df1  # Gets columns 1&2
df_list[[5]] # Gets columns 9&10

Bonus: Make It Flexible #

If you ever have a DataFrame with a different even number of columns, you can make the code adapt automatically instead of hardcoding 29. Use seq() to generate the starting indices of each column pair:

# Generate starting indices for each column pair (1, 3, 5, ...)
col_starts <- seq(1, ncol(LCMS), by = 2)

df_list <- list()
for (i in seq_along(col_starts)) {
  df_list[[i]] <- LCMS[, col_starts[i]:(col_starts[i]+1)]
}

names(df_list) <- paste0("df", seq_along(col_starts))

This way, even if your DataFrame grows or shrinks (as long as column count is even), the code will still split it correctly.

内容的提问来源于stack exchange，提问作者user1878317