R：连接值，而不是变量

这可以用reduction来做.既然你提到了r4ds，我将使用dplyr::coalesce来处理替换NA个值. "NA"是一个字符串文字，而不是R的NA.我猜你指的是后者.

dataset1 <- as.data.frame(matrix(c("1", "1", NA, NA, NA, NA, "A", NA, "B", "B", "A", "B"), ncol = 2))
colnames(dataset1) = c("Patient", "Treatment")
dataset2 <- as.data.frame(matrix(c("1", "1", "2", "4", "3", NA, "A", NA, "B", "B", "A", "B"), ncol = 2))
colnames(dataset2) = c("Patient", "Treatment")
dataset3 <- as.data.frame(matrix(c("1", "1", NA, NA, NA, NA, "A", NA, "B", "B", "A", "B"), ncol = 2))
colnames(dataset3) = c("Patient", "Treatment")
dataset4 <- as.data.frame(matrix(c("1", "1", NA, "2", NA, NA, "A", NA, "B", "B", "A", "B"), ncol = 2))
colnames(dataset4) = c("Patient", "Treatment")
dataset5 <- as.data.frame(matrix(c("1", "1", NA, "2", NA, "2", "A", "C", "B", "B", "A", "B"), ncol = 2))
colnames(dataset5) = c("Patient", "Treatment")

因为你参考了Hadley Wickham的R4DS，我建议使用dplyr::coalesce和purrr::map2_dfc，虽然它们不是必需的，基本R版本也很容易.

现在是解决方案:

Reduce(function(prev, this) purrr::map2_dfc(prev, this, .f = dplyr::coalesce), 
       list(dataset1, dataset2, dataset3, dataset4, dataset5))
# # A tibble: 6 × 2
#   Patient Treatment
#   <chr>   <chr>    
# 1 1       A        
# 2 1       C        
# 3 2       B        
# 4 4       B        
# 5 3       A        
# 6 2       B        

帧的顺序很重要.一旦一个"细胞"是非NA，之后的任何东西都将被忽略(我认为这是意图).例如，如果我们颠倒帧的顺序，我们会得到一个稍微不同的结果(在Patient中):

Reduce(function(prev, this) purrr::map2_dfc(prev, this, .f = coalesce), list(dataset5, dataset4, dataset3, dataset2, dataset1))
# # A tibble: 6 × 2
#   Patient Treatment
#   <chr>   <chr>    
# 1 1       A        
# 2 1       C        
# 3 2       B        
# 4 2       B        
# 5 3       A        
# 6 2       B