如何將缺失的NA包括在不平衡面板數據集的R data.frame中？

Question

我的面板數據不平衡，需要包括所有缺失的觀察值。 例如，我有這樣的事情：

         YEAR    VAR
 FIRM.1  YEAR.1  x.1
 FIRM.1  YEAR.3  x.2
 FIRM.2  YEAR.2  x.3
 FIRM.2  YEAR.3  x.4

我想添加缺少的NA：

         YEAR    VAR
 FIRM.1  YEAR.1  x.1
 FIRM.1  YEAR.2  NA
 FIRM.1  YEAR.3  x.2
 FIRM.2  YEAR.1  NA
 FIRM.2  YEAR.2  x.3
 FIRM.2  YEAR.3  x.4  

如何最方便地做到這一點？

Answer 1

我將使用expand.grid和merge 。

假設您的數據如下：

mydf <- structure(list(FIRM = c("FIRM.1", "FIRM.1", "FIRM.2", "FIRM.2"),
    YEAR = c("YEAR.1", "YEAR.3", "YEAR.2", "YEAR.3"), VAR = c("x.1", "x.2",
    "x.3", "x.4")), .Names = c("FIRM", "YEAR", "VAR"),
    class = "data.frame", row.names = c(NA, -4L))
mydf
#     FIRM   YEAR VAR
# 1 FIRM.1 YEAR.1 x.1
# 2 FIRM.1 YEAR.3 x.2
# 3 FIRM.2 YEAR.2 x.3
# 4 FIRM.2 YEAR.3 x.4

使用expand.grid創建“ FIRM”和“ YEAR”數據的“完整”集，然后merge 。

merge(mydf, expand.grid(FIRM = unique(mydf$FIRM), 
                        YEAR = unique(mydf$YEAR)), 
      all.y = TRUE)
#     FIRM   YEAR  VAR
# 1 FIRM.1 YEAR.1  x.1
# 2 FIRM.1 YEAR.2 <NA>
# 3 FIRM.1 YEAR.3  x.2
# 4 FIRM.2 YEAR.1 <NA>
# 5 FIRM.2 YEAR.2  x.3
# 6 FIRM.2 YEAR.3  x.4