在R语言中合并三个不同列名与行数的数据框的方法

Got it, let's break this down! When you're working with three data frames that have mismatched column names, different row counts, and no row names, the standard merge() function falls short because it expects shared columns to join on. The fix depends on whether you want to stack the data vertically (add all rows together) or bind columns side-by-side (keep all columns, filling missing rows with NA). Here are the two most reliable approaches:

1. Vertical Stack (Combine All Rows) #

If you want to pile all rows from the three data frames into one (with missing columns filled in as NA), use dplyr::bind_rows()—it automatically matches column names and handles mismatches gracefully.

Example Code: #

# Load dplyr (install first if needed: install.packages("dplyr"))
library(dplyr)

# Sample data frames (matching your scenario)
df1 <- data.frame(score = c(85, 92, 78), subject = c("Math", "English", "Science"))
df2 <- data.frame(grade = c("A", "B", "C", "A"), student_id = c(101, 102, 103, 104))
df3 <- data.frame(score = c(90, 81), teacher = c("Ms. Lee", "Mr. Clark"))

# Stack all rows together
combined_vertical <- bind_rows(df1, df2, df3)

This will create a single data frame with all columns (score, subject, grade, student_id, teacher), where any missing values for a row/column pair are filled with NA. All rows from the original three data frames are preserved.

If you prefer base R, you can use plyr::rbind.fill() (though dplyr is more modern and widely used now).

2. Horizontal Bind (Combine All Columns) #

If you want to place the three data frames side-by-side (keeping all columns, filling missing rows with NA), you can't use base R's cbind() directly—it requires identical row counts. Instead, add a shared row ID column to each data frame, then use full_join() to merge them by this ID.

Example Code: #

library(dplyr)

# Add a row ID to each data frame
df1$row_id <- seq(nrow(df1))
df2$row_id <- seq(nrow(df2))
df3$row_id <- seq(nrow(df3))

# Merge all three using full_join (preserves all rows/columns)
combined_horizontal <- full_join(df1, df2, by = "row_id") %>%
  full_join(df3, by = "row_id")

# Optional: Remove the row_id column if you don't need it
combined_horizontal <- select(combined_horizontal, -row_id)

In base R, you can achieve the same with repeated merge() calls using all = TRUE:

# Base R alternative
combined_horizontal <- merge(df1, df2, by = "row_id", all = TRUE)
combined_horizontal <- merge(combined_horizontal, df3, by = "row_id", all = TRUE)
combined_horizontal$row_id <- NULL # Remove row ID

This method ensures every column from all three data frames is included, and rows are aligned by their original position (with NA filling in gaps where one data frame has more rows than others).

Why Your Initial merge() Failed #

The default merge() function looks for shared column names to join on. Since your data frames have no common columns (and no row names to use as a key), it couldn't align the rows/columns properly. Adding a row ID gives it a clear key to merge all data without losing anything.

内容的提问来源于stack exchange，提问作者user44212