在 R 中使用多列按键和年份合并行

Merge rows by key and year with multiple columns in R

我有以下数据集:

df1 <- data.frame(
  "key" = c(1, 1, 1, 1, 2, 2, 2, 2, 3, 3), 
  "year" = c(2002, 2002, 2004, 2004, 2002, 2002, 2004, 2004, 2004, 2004),
  "Var1" = c(10, NA, 5, 5, 4, NA, NA, 3, 2, 2),
  "Var2" = c(1, 1, 3, 3, 2, NA, 3, NA, 1, NA),
  "Var3" = c(NA, 2, NA, NA, 5, 5, 3, NA, 2, NA),
  "Var4" = c(NA, 4, 5, 5, 6, NA, 4, NA, NA, NA))

我现在想按键和年份合并重复的行以获得如下所示的数据集:

df2 <- data.frame(
  "key" = c(1, 1, 2, 2, 3), 
  "year" = c(2002, 2004, 2002, 2004, 2004),
  "Var1" = c(10, 5, 4, 3, 2),
  "Var2" = c(1, 3, 2, 3, 1),
  "Var3" = c(2, NA, 5, 3, 2),
  "Var4" = c(4, 5, 6, 4, NA))

问题是我有超过 30 列和成百上千行。因此,这个解决方案似乎有点不方便:Merge rows within a dataframe by a key。 如果有任何帮助,我将不胜感激!

您可以group_by(key, year)并获取每列的最大值,不包括 NA 和只有 NA 的组:

library(dplyr)
df1 %>% 
  group_by(key, year) %>% 
  summarise(across(everything(), ~ ifelse(all(is.na(.x)), NA, max(.x, na.rm = T))))

## A tibble: 5 x 6
## Groups:   key [3]
#    key  year  Var1  Var2  Var3  Var4
#  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#1     1  2002    10     1     2     4
#2     1  2004     5     3    NA     5
#3     2  2002     4     2     5     6
#4     2  2004     3     3     3     4
#5     3  2004     2     1     2    NA

您可以使用 fill() 按每个组填充缺失值,并使用 distinct() 查找唯一行。

library(tidyverse)

df1 %>%
  group_by(key, year) %>%
  fill(Var1:Var4, .direction = "downup") %>%
  distinct() %>%
  ungroup()

# A tibble: 5 × 6
    key  year  Var1  Var2  Var3  Var4
  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1     1  2002    10     1     2     4
2     1  2004     5     3    NA     5
3     2  2002     4     2     5     6
4     2  2004     3     3     3     4
5     3  2004     2     1     2    NA