Question

我想使用第三个data.frame matches合并两个data.frame，它提供了如何合并前两个数据的键（id和number是匹配的键。Class应该始终匹配，并且用作附加检查）。 df2应该合并到df1中，以替换具有相同名称的列，并将变量df2中的变量插入到df1的相应行中（在此示例中，这些行将被计分，但在我的实际示例中，大约20列以上的列应在30列左右的列中进行替换（其余10列将保持不变）。我希望这一点很清楚，并在此先感谢非常感谢，以帮助您高效地进行操作。由于我想将所有记录保留在left_join中，因此它必须为df1。 dplyr是首选项！

df1 <- data.frame(1:10,
                  c("a","b","a","b","c","c","b","a","b","c"),
                  seq(20,38,2),
                  dnorm(10,5,2),
                  rnorm(10,78,5))
colnames(df1) <- c("id","class","score","misc","hc")
df2 <- data.frame(11:19,
                  c("g","d","a","e","c","d","f","a","b"),
                  seq(40,48,1),
                  dnorm(9,6,1),
                  rnorm(9,45,3))
colnames(df2) <- c("number","class","score","other","md")

matches <- data.frame(c("b","a","c","a"),
                      c(4,1,5,8),
                      c(19,18,15,13))
colnames(matches) <- c("class","id","number")

library(dplyr)

预期输出：

id class          misc       hc number   score       other       md
 1     a    0.00876415 70.85210     18      47 0.004431848 40.99189
 2     b    0.00876415 81.27802     22      22          NA       NA
 3     a    0.00876415 69.19300     24      24          NA       NA
 4     b    0.00876415 75.86247     19      48 0.004431848 44.78242
 5     c    0.00876415 69.18268     15      44 0.004431848 42.70708
 6     c    0.00876415 80.49475     30      30          NA       NA
 7     b    0.00876415 76.52195     32      32          NA       NA
 8     a    0.00876415 73.52712     13      42 0.004431848 45.69842
 9     b    0.00876415 81.50631     36      36          NA       NA
10     c    0.00876415 80.55145     38      38          NA       NA

Answer 1

使用dplyr

library(dplyr) 
df1 %>% left_join(matches,by=c('id','class'),suffix=c('.df1','.df2')) %>%
        left_join(df2,by=c('class','number'),suffix=c('.df1','.df2')) %>% 
        mutate(score=if_else(!is.na(score.df2),score.df2,score.df1)) %>% 
        select(-score.df1,score.df2)

Answer 2

只做两个这样的合并：

tmp = merge(df1, matches, all.x = FALSE, all.y = TRUE)
res = merge(tmp, df2, all.x = TRUE, all.y = FALSE)
res

使用辅助data.frame dplyr R合并两个data.frame

2 个答案: