Question

我是R编程的初学者，希望获得有关在数据集上执行日期运算的帮助（希望可以在R中做到这一点），

我已将给定的数据集导入为“交易”

deals <- read.csv("deals_dates.csv")

带有数据（输入），

Deal_Id  Deal_Name       Start_Date   End_Date
1        Samsung Mobile  3/5/2018     6/23/2018
2        Apple Watch     12/21/2017   2/9/2018

我想找到每个交易在给定开始日期/结束日期范围之间的每月开始和结束日期，

（预期输出）

Deal_Id  Deal_Name       Start_Date   End_Date
1        Samsung Mobile  3/5/2018     3/31/2018
1        Samsung Mobile  4/1/2018     4/30/2018
1        Samsung Mobile  5/1/2018     5/31/2018
1        Samsung Mobile  6/1/2018     6/23/2018
2        Apple Watch     12/21/2017   12/31/2017
2        Apple Watch     1/1/2018     1/31/2018
2        Apple Watch     2/1/2018     2/9/2018

谢谢！

Answer 1

您还可以仅使用Vanilla R来执行某些操作。例如，具有以下功能：

myDateFunction <-function(dealName,stdate,endate){

 firstDay <- cut(as.Date(stdate,format = "%m/%d/%Y"), "month")
 Start <-seq(as.Date(firstDay), as.Date(endate,format = "%m/%d/%Y"), by = "month")
 DealName <- rep(dealName,length(Start))
 plusOne<-seq(as.Date(firstDay), by = "month", length = 2)[2]
 End<-seq(as.Date(plusOne),length=4,by="months")-1

 data <- data.frame(DealName,Start,End)
 data$Start[1] <- as.Date(stdate,format = "%m/%d/%Y")
 data$End[-1][3] <- as.Date(endate,format = "%m/%d/%Y")
 return(data)

}

如果我们这样称呼：

myDateFunction("Samsung Mobile","3/5/2018","6/23/2018")

产生以下输出：

    DealName      Start        End
1 Samsung Mobile 2018-03-05 2018-03-31
2 Samsung Mobile 2018-04-01 2018-04-30
3 Samsung Mobile 2018-05-01 2018-05-31
4 Samsung Mobile 2018-06-01 2018-06-23

Answer 2

我们可以先将“日期”列转换为Date类，然后使用seq按“月”获取“日期”序列，floor将“日期”，和unnest

library(tidyverse)
df1 %>%
   mutate_at(vars(ends_with("Date")), mdy) %>%
   group_by(Deal_Id, Deal_Name) %>% 
   nest %>% 
   mutate(data = map(data, ~ 
       seq(.x$Start_Date, .x$End_Date, by = "1 month") %>%
       {x1 <- floor_date(., unit = "month")
        x2 <-  (x1[-1]-1)
       tibble(Start_Date = c(.x$Start_Date[1], x1[-1]),
              End_date = c(x2, last(.x$End_Date)))})) %>% 
   unnest

Answer 3

data.table解决方案，并使用tidyverse作了一些准备。实际的周期修补由data.table的非常快的foverlaps()完成。如果您使用的是大数据集，请尝试一下此解决方案。

library( tidyverse )
library( data.table )

#create df with start and end dates
dt1 <- data.table( start = seq( as.Date("2017-01-01"), 
                               length.out = 24,
                               by = "month") ,
                  end = seq( as.Date("2017-02-01"), 
                               length.out = 24,
                               by = "month" ) - 1,
                  stringsAsFactors = FALSE)

#read data with deals
df2 <- read.table(text=" Deal_Id  Deal_Name       Start_Date   End_Date
1        Samsung_Mobile  3/5/2018     6/23/2018
2        Apple_Watch     12/21/2017   2/9/2018", header = T)

#set the dates as actual dates
dt2 <- df2 %>%
  mutate( start = as.Date( as.character(Start_Date), format = "%m/%d/%Y"),
          end = as.Date( as.character(End_Date), format = "%m/%d/%Y") ) %>%
  select( -Start_Date, -End_Date) %>%
  setDT()

setkey(dt1, start, end)
setkey(dt2, start, end)

#create overlap join
dt3 <- foverlaps( dt1, dt2, type = "any", nomatch = 0L)
#set start 
dt3[start < i.start, start := i.start]
#set end
dt3[end > i.end, end := i.end]

#print results
dt3[, c("i.end", "i.start") := NULL ][]

#    Deal_Id      Deal_Name      start        end
# 1:       2    Apple_Watch 2017-12-21 2017-12-31
# 2:       2    Apple_Watch 2018-01-01 2018-01-31
# 3:       2    Apple_Watch 2018-02-01 2018-02-09
# 4:       1 Samsung_Mobile 2018-03-05 2018-03-31
# 5:       1 Samsung_Mobile 2018-04-01 2018-04-30
# 6:       1 Samsung_Mobile 2018-05-01 2018-05-31
# 7:       1 Samsung_Mobile 2018-06-01 2018-06-23

在R中给定日期范围之间查找每个月的开始和结束日期

3 个答案: