根据两列R中的匹配行填充列中的空白

Question

In the df2 I would like to fill in gaps in column d based on matching records in columns b and c between two dataframes. What would be a quick and elegant way to do that? It is important to mention it should work well for occasions where matching rows might have different locations in both dataframes.

df1 <- data.frame( a = c(1,1,1,1,1,2,2,2,2,2) ,b = rep(seq(41,45,1),each=2), c = c(101:105,101:105), d = LETTERS[seq( from = 1, to = 10 )])
df2 <- data.frame( a = c(1,1,1,1,1,2,2,2,2,2) ,b = rep(seq(41,45,1),each=2), c = c(101:105,101:105), d = c(LETTERS[seq( from = 1, to = 6 )],rep(NA,4)))

> df1
   a  b   c d
1  1 41 101 A
2  1 41 102 B
3  1 42 103 C
4  1 42 104 D
5  1 43 105 E
6  2 43 101 F
7  2 44 102 G
8  2 44 103 H
9  2 45 104 I
10 2 45 105 J
> df2
   a  b   c    d
1  1 41 101    A
2  1 41 102    B
3  1 42 103    C
4  1 42 104    D
5  1 43 105    E
6  2 43 101    F
7  2 44 102 <NA>
8  2 44 103 <NA>
9  2 45 104 <NA>
10 2 45 105 <NA>

The result should be following:

   a  b   c d
1  1 41 101 A
2  1 41 102 B
3  1 42 103 C
4  1 42 104 D
5  1 43 105 E
6  2 43 101 F
7  2 44 102 G
8  2 44 103 H
9  2 45 104 I
10 2 45 105 J

Answer 1

While you can do lookups with match and perhaps %in% , I'd think another (robust) way to do it is with a merge/join:

df2mod <- merge(df2, df1[,c('b','c','d')], by = c("b", "c"), all=TRUE)
df2mod
#     b   c a  d.x d.y
# 1  41 101 1    A   A
# 2  41 102 1    B   B
# 3  42 103 1    C   C
# 4  42 104 1    D   D
# 5  43 101 2    F   F
# 6  43 105 1    E   E
# 7  44 102 2 <NA>   G
# 8  44 103 2 <NA>   H
# 9  45 104 2 <NA>   I
# 10 45 105 2 <NA>   J

In this case, dx are the original df2$d . Because your data is factor s, some extra parts are necessary ( as.character and the re factor ).

df2mod$d <- with(df2mod, ifelse(is.na(d.x), as.character(d.y), as.character(d.x)))
df2mod$d <- factor(df2mod$d, levels = levels(df1$d))
df2mod
#     b   c a  d.x d.y d
# 1  41 101 1    A   A A
# 2  41 102 1    B   B B
# 3  42 103 1    C   C C
# 4  42 104 1    D   D D
# 5  43 101 2    F   F F
# 6  43 105 1    E   E E
# 7  44 102 2 <NA>   G G
# 8  44 103 2 <NA>   H H
# 9  45 104 2 <NA>   I I
# 10 45 105 2 <NA>   J J
df2mod[,c("d.x", "d.y")] <- NULL # cleanup unnecessary columns