如何重新排列数据框，从两个变量和第三个具有两个级别的分类变量中创建四个连续变量？

Question

我有下一个数据框：

df <- data.frame(Date= as.Date(c("2020-08-04","2020-08-04","2020-08-06","2020-08-06","2020-08-07","2020-08-07")),
                   Period= c("Day","Night","Day","Night","Day","Night"),
                   State.1= c(1, 0.45,0.48,0.32,0.29,0.87),
                   State.2= c(0, 0.55,0.28,0.62,0.79,0.17))
  
  df

         Date Period State.1 State.2
1 2020-08-04    Day    1.00    0.00
2 2020-08-04  Night    0.45    0.55
3 2020-08-06    Day    0.48    0.28
4 2020-08-06  Night    0.32    0.62
5 2020-08-07    Day    0.29    0.79
6 2020-08-07  Night    0.87    0.17

我在其中根据 date 和 Period 指出了 State.1 和 State.2 的一些值。出于说明目的，我想重新安排我的数据框，将 Period 包含在 State.1 和 State.2 中。我希望得到下一个：

df2

       Date State.1_day State.1_night State.2_day State.2_night
1 2020-08-04        1.00          0.45        0.00          0.55
2 2020-08-05          NA            NA          NA            NA
3 2020-08-06        0.48          0.32        0.28          0.62
4 2020-08-07        0.29          0.87        0.79          0.17

我该怎么做？我尝试了 melt() 但我无法得到我想要的。

提前致谢。

Answer 1

你可以这样做：

tidyr::pivot_wider(df, names_from = Period, values_from = c("State.1", State.2))
#> # A tibble: 3 x 5
#>   Date       State.1_Day State.1_Night State.2_Day State.2_Night
#>   <date>           <dbl>         <dbl>       <dbl>         <dbl>
#> 1 2020-08-04       1              0.45        0             0.55
#> 2 2020-08-06       0.48           0.32        0.28          0.62
#> 3 2020-08-07       0.290          0.87        0.79          0.17

或者如果您想要每一天的记录，即使原始数据框中不存在这一天，您可以这样做：

dplyr::left_join(data.frame(Date = seq(min(df$Date), max(df$Date), by = "day")),
                 tidyr::pivot_wider(df, names_from = Period, 
                                   values_from = c("State.1", State.2)))

#>         Date State.1_Day State.1_Night State.2_Day State.2_Night
#> 1 2020-08-04        1.00          0.45        0.00          0.55
#> 2 2020-08-05          NA            NA          NA            NA
#> 3 2020-08-06        0.48          0.32        0.28          0.62
#> 4 2020-08-07        0.29          0.87        0.79          0.17

Answer 2

我们可以使用data.table

library(data.table)
dcast(setDT(df), Date ~ Period, value.var = c("State.1", "State.2"))

如何重新排列数据框，从两个变量和第三个具有两个级别的分类变量中创建四个连续变量？

How to re-arrange a dataframe creating four continuous variables from two ones and a third cathegorical variable with two levels?

r

melt