使用 dplyr 填充缺失值（通过连接？）

我有一个数据框（df1），其中缺少一些值（城市、州）：

SiteID    City          StateBasedIn   Lat  Lon   Var1 Var2
4227      Richmond      KY            -39  -113   6    0
4987      Nashville     TN            -33  -97    7    0
4000      Newark        NJ            -39  -95    8    0
4925      Miami         FL            -40  -99    0    0
4437      Montgomery    AL            -32  -117   4    1
4053      Jonesboro     AR            -30  -98    8    1

df1 <- structure(list(SiteID = c(4227L, 4987L, 4000L, 4925L, 4437L, 
4053L, 4482L, 4037L, 4020L, 1787L, 2805L, 3025L, 3027L, 3028L, 
3029L, 3030L, 3031L, 3033L), City = structure(c(10L, 7L, 8L, 
5L, 6L, 4L, 2L, 9L, 3L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = c("", 
"Arcata", "Jackson", "Jonesboro", "Miami", "Montgomery", "Nashville", 
"Newark", "Portland", "Richmond"), class = "factor"), StateBasedIn = structure(c(6L, 
10L, 8L, 5L, 2L, 3L, 4L, 9L, 7L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L), .Label = c("", "AL", "AR", "CA", "FL", "KY", "MS", "NJ", 
"OR", "TN"), class = "factor"), Lat = c(-39L, -33L, -39L, -40L, 
-32L, -30L, -38L, -31L, -35L, -38L, -30L, -39L, -38L, -32L, -39L, 
-31L, -38L, -34L), Lon = c(-113L, -97L, -95L, -99L, -117L, -98L, 
-98L, -95L, -112L, -120L, -114L, -81L, -117L, -90L, -109L, -115L, 
-81L, -104L), Var1 = c(6L, 7L, 8L, 0L, 4L, 8L, 1L, 8L, 0L, 3L, 
3L, 7L, 4L, 8L, 0L, 8L, 1L, 3L), Var2 = c(0L, 0L, 0L, 0L, 1L, 
1L, 1L, 1L, 1L, 0L, 1L, 1L, 0L, 0L, 1L, 0L, 1L, 0L)), .Names = c("SiteID", 
"City", "StateBasedIn", "Lat", "Lon", "Var1", "Var2"), class = "data.frame", row.names = c(NA, 
-18L))

我想通过与另一个数据框 (df2) 合并来填充这些值，该数据框具有 3 个相同的列，但不是 df1 中的所有列：

SiteID       City         StateBasedIn
1787         Lusby        MD
2805         Springdale   AR
3025         Saukville    WI
3027         Saukville    WI
3028         Saukville    WI
3029         Saukville    WI

df2 <- structure(list(SiteID = c(1787L, 2805L, 3025L, 3027L, 3028L, 
3029L, 3030L, 3031L, 3033L), City = structure(c("Lusby", "Springdale", 
"Saukville", "Saukville", "Saukville", "Saukville", "Saukville", 
"Mequon", "Mequon"), .Dim = c(9L, 1L)), StateBasedIn = structure(c("MD", 
"AR", "WI", "WI", "WI", "WI", "WI", "WI", "WI"), .Dim = c(9L, 
1L))), row.names = c(NA, -9L), class = "data.frame", .Names = c("SiteID", 
"City", "StateBasedIn"))

所以基本上我会保留 df1 中的所有信息，并输入 df2 中可用的缺失值。由于我还不太熟悉所有 dplyr 选项，因此我尝试了不同的“加入”选项，但没有成功。我还尝试在基础包中使用“合并”，但仍然没有成功。还有另一种方法可以做到这一点（最好使用 dplyr）？

您可以使用full_join from dplyr， 随着replace and coalesce整理出一个非常简洁的解决方案。

library(dplyr)
library(purrr) 

# Cleaning from r2evans (if you want to keep it to dplyr just use r2evans lapply method

df1 <- mutate_if(df1, is.factor, as.character)
df2 <- dmap(df2, as.vector)

full_join(df1, df2, by = "SiteID") %>% 
  mutate_at(vars(matches("City","StateBased")), funs(replace(., . == "", NA))) %>% 
  mutate(City = coalesce(City.y, City.x),
         StateBasedIn = coalesce(StateBasedIn.y, StateBasedIn.x)) %>% 
  select(-contains("."))