在R中对DataFrame进行子集筛选：保留列首元素并移除空行的实现方法

Great question! Let's work through this with clean, concise R code that matches your requirements. First, let's confirm the task: we need to retain only the first occurrence (topmost non-empty value) in each category1 to category5 column, then remove any rows where all these category columns are completely empty.

First, let's start with your original DataFrame for reference:

df <- data.frame( 
  Hits = c("Hit1", "Hit2", "Hit3", "Hit4", "Hit5"),
  category1 = c("a1", "", "b1", "a1", "c1"),
  category2 = c("", "", "", "", "a2"),
  category3 = c("a3", "", "b3", "", "a3"),
  category4 = c("", "", "", "", ""),
  category5 = c("", "", "a5", "b5", ""),
  stringsAsFactors = FALSE
)

Solution 1: Keep only the very first non-empty value per category column #

If your goal is to keep only the topmost non-empty entry in each category column (setting all other values in that column to empty), here's a streamlined solution using dplyr:

library(dplyr)

df_processed <- df %>%
  # Target all category columns, retain only the first non-empty value
  mutate(across(starts_with("category"), 
                ~ ifelse(row_number() == which(. != "")[1], ., ""))) %>%
  # Drop rows where every category column is empty
  filter(if_any(starts_with("category"), ~ . != ""))

Let's break this down:

• across(starts_with("category"), ...) lets us apply the same logic to all category columns in one go
• which(. != "")[1] finds the row number of the first non-empty value in the column; we keep that value and set everything else to an empty string
• filter(if_any(...)) removes rows where all category columns are empty (that's row 2 in your original data)

The resulting df_processed will look like this:

Hits category1 category2 category3 category4 category5
1 Hit1        a1                    a3                  
3 Hit3                             a5                  
5 Hit5                             a2                  

Solution 2: Keep the first occurrence of each unique value per category column #

If you actually meant to retain the first instance of every unique non-empty value in each column (and clear out any duplicate values that appear later), use this variation:

df_processed <- df %>%
  # Clear out duplicate non-empty values in each category column
  mutate(across(starts_with("category"), 
                ~ ifelse(duplicated(.) & . != "", "", .))) %>%
  # Remove rows with all empty category columns
  filter(if_any(starts_with("category"), ~ . != ""))

Here, duplicated(.) & . != "" identifies non-empty values that have appeared earlier in the column; we set those to empty. The output here would be:

Hits category1 category2 category3 category4 category5
1 Hit1        a1                    a3                  
3 Hit3        b1                    b3                   a5
4 Hit4                             b5                  
5 Hit5        c1        a2                              

Both solutions are efficient and readable, using dplyr's vectorized operations to avoid messy loops. Pick the one that aligns with your exact needs!

内容的提问来源于stack exchange，提问作者accibio