如何通过循环批量导入并标准化处理多份月度新闻DataFrame？

Absolutely! You don’t have to repeat code 264 times—we can wrap your single-month logic into a reusable function and loop through all your target months. Here’s how to do it step by step:

Step 1: Generate the Sequence of Months to Process #

First, create a list of all year-month combinations from January 1996 to December 2018. We’ll format them as YYYYMM strings to match your file naming convention:

# Create a sequence of dates from 1996-01 to 2018-12
month_dates <- seq.Date(from = as.Date("1996-01-01"), 
                        to = as.Date("2018-12-01"), 
                        by = "month")

# Convert dates to YYYYMM strings (e.g., "199601", "199602")
ym_list <- format(month_dates, "%Y%m")

Step 2: Wrap Your Single-Month Logic into a Function #

Turn your existing processing steps into a function that accepts a YYYYMM string as input. This keeps your code clean and reusable:

Note: I’m assuming you’ve already defined tags_split (your list of tags) and rx.app (your text matching rule) elsewhere in your script—make sure these are loaded before running the function.

process_monthly_news <- function(ym) {
  # Extract year from YYYYMM string for file path
  year <- substr(ym, 1, 4)
  
  # Build file path dynamically
  file_path <- paste0("D:/Reuters/", year, "/News.RTRS.", ym, ".0210.txt.gz")
  
  # Import data
  news_data <- read.delim(file_path, header = FALSE, quote = "")
  
  # Split first column into additional variables
  news_data <- news_data %>%
    mutate(v2 = lapply(strsplit(as.character(V1), "\"mimeType\""), "[", 2))
  
  # Filter news containing "R:"
  news_data <- news_data[grepl('R:', news_data$v2), ]
  
  # Filter news matching your tags and combine results
  tag_dfs <- list()
  for(i in 1:length(tags_split)) {
    tags_split1 <- paste(tags_split[[i]], collapse = "|")
    tags_split1 <- gsub("[[:space:]]", "", tags_split1)
    tag_dfs[[i]] <- news_data[grepl(tags_split1, news_data$v2, perl = TRUE), ]
  }
  news_data <- do.call(rbind, tag_dfs)
  
  # Remove duplicates
  news_data <- news_data[!duplicated(news_data), ]
  
  # Text analysis with rx.app
  news_data <- news_data %>%
    mutate(approach = lengths(regmatches(v2, gregexpr(rx.app, v2, perl = TRUE))))
  
  # Export processed data to CSV
  output_file <- paste0("news.", ym, ".csv")
  write.csv(news_data, file = output_file, row.names = FALSE)
  
  # Clean up temporary variables to save memory
  rm(news_data, tag_dfs)
  gc() # Optional: Force garbage collection
}

Step 3: Run the Loop to Process All Months #

Now iterate over every YYYYMM string in ym_list and call your function:

Option 1: For Loop (Great for Tracking Progress) #

for(ym in ym_list) {
  cat("Processing", ym, "...\n")
  process_monthly_news(ym)
}

Option 2: lapply (More Concise) #

lapply(ym_list, process_monthly_news)

Optional: Add Error Handling #

If some files might be missing or corrupted, wrap the function call in tryCatch to avoid breaking the entire loop:

for(ym in ym_list) {
  cat("Processing", ym, "...\n")
  tryCatch({
    process_monthly_news(ym)
  }, error = function(e) {
    cat("Failed to process", ym, ":", e$message, "\n")
  })
}

This way, if one file has an issue, the loop will continue processing the rest.

内容的提问来源于stack exchange，提问作者Esperanta