合并多个数据帧的列，根据列名覆盖值

Question

I have the following data:

x<-data.frame(A=c(1,NA),B=c(NA,NA),C=c(NA,1))
   A  B  C
1  1 NA NA
2  NA NA  1

and

y<-data.frame(A=c(NA,0),C=c(NA,NA),D=c(NA,0))
   A  C  D
1 NA NA NA
2  0 NA  0

and I want to merge them in such a way, that columns of the same name are overwritten giving precedence to non - NA values. In addition, columns that are not common should be added.

Desired result:

  A  B  C  D
1 1 NA NA NA
2 0 NA  1  0

There are no conflicts between non - NA values in my data.

Answer 1

you can try:

library(dplyr)
library(magrittr)

full_join(x,y) %>%
      inner_join(y) %>%
      coalesce(full_join(x,y) %>%
                 inner_join(x))

which gives:

  A  B  C  D
1 1 NA NA NA
2 0 NA  1  0

Answer 2

Here is a base R method that should be fairly general.

# fill in x NA values with values in y
x[] <- lapply(names(x), function(i) ifelse(is.na(x[,i]) & i %in% names(y), y[,i], x[,i]))

# add variables in y that are not in x and provide names to the variables
setNames(cbind(x, y[, !(names(y) %in% names(x))]), union(names(x), names(y)))
  A  B  C  D
1 1 NA NA NA
2 0 NA  1  0