如何在R中逐行传递函数并将结果导出为数据框

Hey there! Let's fix up your approach to this problem—you're right that the current loop has a couple of issues, and we can make this cleaner and more effective with R's functional programming tools.

First, let's spot the problems in your existing loop:

• You're passing the entire df2 to do_reports every time instead of the single row df2[i,]
• You're overwriting create_report in each iteration, so you'll only end up with the last run's results instead of saving all of them

Here are a couple of better approaches, starting with base R (no extra packages needed) and then a more concise option with the purrr package:

Base R Solution #

We'll use lapply to iterate over each row index of df2, run your function for each row, collect all results in a list, then combine them into final dataframes:

# Step 1: Run do_reports for each row of df2 and collect all results
report_results <- lapply(seq_len(nrow(df2)), function(row_idx) {
  do_reports(df1, df2[row_idx, ], df3)
})

# Step 2: Combine all create_report[1] results into one dataframe
combined_report1 <- do.call(rbind, lapply(report_results, function(x) x[[1]]))

# Step 3: Combine all create_report[2] results into one dataframe
combined_report2 <- do.call(rbind, lapply(report_results, function(x) x[[2]]))

Cleaner Solution with purrr (tidyverse) #

If you're open to using the purrr package (part of the tidyverse, great for iterative tasks), this code is more readable and concise:

# Install/load purrr if you haven't already
# install.packages("purrr")
library(purrr)

# Collect all function outputs
report_results <- map(seq_len(nrow(df2)), ~do_reports(df1, df2[., ], df3))

# Extract and row-bind the first report type
combined_report1 <- map_dfr(report_results, ~.[[1]])

# Extract and row-bind the second report type
combined_report2 <- map_dfr(report_results, ~.[[2]])

Bonus: Add a Row Identifier (Optional) #

If you want to track which row of df2 each result came from, you can add an extra column to each output before combining:

# Using base R
report_results <- lapply(seq_len(nrow(df2)), function(row_idx) {
  res <- do_reports(df1, df2[row_idx, ], df3)
  # Add a column to track the df2 row number
  res[[1]]$df2_source_row <- row_idx
  res[[2]]$df2_source_row <- row_idx
  res
})

# Then combine as before
combined_report1 <- do.call(rbind, lapply(report_results, function(x) x[[1]]))

This way, you can easily trace back any entries in your final reports to the specific row of df2 that generated them.

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