R：取消堆叠带有日期的列

Question

假设我有以下数据框：

df <- data.frame(Order=c("1234567","1234567","1234567","456789","456789"),Stage=c("Pipeline","Proposal","Closed","Pipeline","Lost"),StageChange=c("2008-01-01","2008-01-02","2008-01-03","2008-01-10","2008-01-12"))

导致：

    head(df)
    Order    Stage StageChange
1 1234567 Pipeline  2008-01-01
2 1234567 Proposal  2008-01-02
3 1234567   Closed  2008-01-03
4  456789 Pipeline  2008-01-10
5  456789     Lost  2008-01-12

我需要拆开 "Stage" 列并得到这样的数据框：

    Order   Pipeline   Proposal     Closed       Lost
1 1234567 2008-01-01 2008-01-02 2008-01-03         NA
2  456789 2008-01-10         NA         NA 2008-01-12

我阅读了文档并使用 dplyr 和 tidyr () 尝试了不同的方法，但我的无知是成功的。

有什么想法可以完成我需要的吗？

我的objective，说白了，就是用这个数据来计算一个特定的Order在特定的Stage上停留的天数。有些订单丢失，有些订单已关闭（赢得），这就是为什么有 "NA" 个值的原因。当订单没有更改到特定阶段时也会发生同样的情况（订单可以从管道转到丢失，中间阶段没有任何更改）。

谢谢！

Answer 1

您可能会使用 tidyr::pivot_wider。那是新版本的 retired-function spread

# install.packages("tidyr")
library(tidyr)

df %>%
  pivot_wider(names_from = Stage, values_from = StageChange)

# # A tibble: 2 x 5
#   Order   Pipeline   Proposal   Closed     Lost      
#   <fct>   <fct>      <fct>      <fct>      <fct>     
# 1 1234567 2008-01-01 2008-01-02 2008-01-03 NA        
# 2 456789  2008-01-10 NA         NA         2008-01-12

Answer 2

使用dplyr::spread

library(dplyr)

df %>% 
  spread(Stage,StageChange) %>% 
  select(Order,Pipeline,Proposal,Closed,Lost)

Answer 3

日期将是 factor class

library(tidyverse)

df_wide <- df %>%
  tidyr::pivot_wider(names_from = Stage, values_from = StageChange)

df_wide

# A tibble: 2 x 5
  Order   Pipeline   Proposal   Closed     Lost      
  <fct>   <fct>      <fct>      <fct>      <fct>     
1 1234567 2008-01-01 2008-01-02 2008-01-03 NA        
2 456789  2008-01-10 NA         NA         2008-01-12

如果您想将日期转换为 Date class

df_wide_dates <- df %>%
  tidyr::pivot_wider(names_from = Stage, values_from = StageChange) %>%
  dplyr::mutate_at(., vars(Pipeline, Proposal, Closed, Lost), as.Date)

df_wide_dates

# A tibble: 2 x 5
  Order   Pipeline   Proposal   Closed     Lost      
  <fct>   <date>     <date>     <date>     <date>    
1 1234567 2008-01-01 2008-01-02 2008-01-03 NA        
2 456789  2008-01-10 NA         NA         2008-01-12

R：取消堆叠带有日期的列

R: Unstack a column with dates

transpose

r

tidyr