转置数据框并将列名添加为 R 中的变量

Question

我有一个时间序列数据框，它很宽并且以个股数据作为列名。我想将此数据框转换为长格式，同时不取消查看数据属于哪个代码的能力。

下面是数据。

df = structure(list(Date = structure(c(1607922000, 1608008400), class = c("POSIXct", 
"POSIXt"), tzone = ""), AAPL.Close = c(0.32982465, 0.34001608
), MSFT.Close = c(0.26307234, 0.27235893), GS.Close = c(0.30742572, 
0.29825025), QQQ.Close = c(0.25350002, 0.24456267)), row.names = 1:2, class = "data.frame")


  Date        AAPL.Close MSFT.Close GS.Close QQQ.Close
1 2020-12-14  0.3298246  0.2630723 0.3074257 0.2535000
2 2020-12-15  0.3400161  0.2723589 0.2982502 0.2445627

我希望新数据框看起来像这样。

Date          Data     Ticker
2020-12-14   .3298     AAPL 
2020-12-15   .3400     AAPL
2020-12-14   .260      MSFT
2020-12-15   .27       MSFT
.
.

感谢您的帮助

Answer 1

我们可以使用 pivot_longer

library(tidyr)
library(dplyr)
df %>% 
   mutate(Date = as.Date(Date)) %>%
   pivot_longer(cols = -Date, names_to = c("Ticker", ".value"), 
       names_sep = "\.") %>% 
   rename(Data = Close)

Answer 2

在过去的几天里，我多次尝试 data.table 尝试额外的 akrun 解决方案:-)

library(data.table)
library(stringr)
setDT(df)

df = melt(df, id.vars=c("Date"), variable.name = "Data", value.name="Ticker")
df[, Data:=str_replace(Data, ".Close", "")]

输出：

> df
                  Date Data    Ticker
1: 2020-12-14 05:00:00 AAPL 0.3298246
2: 2020-12-15 05:00:00 AAPL 0.3400161
3: 2020-12-14 05:00:00 MSFT 0.2630723
4: 2020-12-15 05:00:00 MSFT 0.2723589
5: 2020-12-14 05:00:00   GS 0.3074257
6: 2020-12-15 05:00:00   GS 0.2982502
7: 2020-12-14 05:00:00  QQQ 0.2535000
8: 2020-12-15 05:00:00  QQQ 0.2445627

更新，如果您只想对日期而不是日期时间进行分组：

df = melt(df[, Date:=as.Date(Date)], id.vars=c("Date"), variable.name = "Data", value.name="Ticker")

Answer 3

library(data.table)
library(magrittr)

setDT(df)

melt(data = df, id.vars = "Date") %>% 
  .[, variable := gsub(pattern = "(.+)\.(.+)", replacement = "\1", x = variable)] %>% 
  .[]


                  Date variable     value
1: 2020-12-14 08:00:00     AAPL 0.3298246
2: 2020-12-15 08:00:00     AAPL 0.3400161
3: 2020-12-14 08:00:00     MSFT 0.2630723
4: 2020-12-15 08:00:00     MSFT 0.2723589
5: 2020-12-14 08:00:00       GS 0.3074257
6: 2020-12-15 08:00:00       GS 0.2982502
7: 2020-12-14 08:00:00      QQQ 0.2535000
8: 2020-12-15 08:00:00      QQQ 0.2445627

转置数据框并将列名添加为 R 中的变量

Transpose a data frame and add column names as variables in R

r

dplyr

data.table

tidyverse