填充日期並在數據框中創建新行

rptdate  st 
1 2/18/2017 2/12/2017 
2 2/25/2017 2/19/2017 
3 3/4/2017 2/26/2017 
4 3/11/2017 3/5/2017 
5 3/18/2017 3/12/2017 
6 3/25/2017 3/19/2017 
7 4/1/2017 3/26/2017 
8 4/8/2017 4/2/2017 
9 4/15/2017 4/9/2017 
10 4/22/2017 4/16/2017 
11 4/29/2017 4/23/2017 
12 5/6/2017 4/30/2017 
13 5/13/2017 5/7/2017 
14 5/20/2017 5/14/2017 
15 5/27/2017 5/21/2017 
16 6/3/2017 5/28/2017 
17 6/10/2017 6/4/2017

所以基本上rptdate是一堆星期六，st是前一個星期天。填充日期並在數據框中創建新行

我想重塑這個數據框（該數據在日期格式）以這種方式：

什麼，我想這樣做是這樣的：

i=1 
j=1 
While (rptdate[i][j]>=st[i][j]) 
    {add a new row where rptdate[i][j+1]= rptdate[i][j] and st[i][j+1]=rptdate[i][j]+1}

所以基本上，我需要的新的數據幀應該是這樣的：

 rptdate  st 
1 2/18/2017 2/12/2017 
    2/18/2017 2/13/2017 
    2/18/2017 2/14/2017 
    2/18/2017 2/15/2017 
    2/18/2017 2/16/2017 
    2/18/2017 2/17/2017 
    2/18/2017 2/18/2017 
2 2/25/2017 2/19/2017 
    2/25/2017 2/20/2017 
    2/25/2017 2/21/2017 
    2/25/2017 2/22/2017 
    2/25/2017 2/23/2017 
    2/25/2017 2/24/2017 
    2/25/2017 2/25/2017

非常感謝您的時間。

來源

2017-06-19 Omi Slash

你能用文字解釋你想要做什麼嗎？ – mtoto

這是一個基於R的想法。您需要先將變量轉換爲日期。然後爲每個日期擴展數據框額外7行（1周）。使用seq生成所有缺失日期，並將其添加到您的st變量中。

d2[] <- lapply(d2, function(i) as.Date(i, format = '%m/%d/%Y')) 
d3 <- d2[rep(row.names(d2), each = 7),] 
d3$st<- do.call(c, Map(function(x, y)seq(x, y, by = 1), d2$st, d2$rptdate)) 

head(d3, 10) 
#  rptdate   st 
#1 2017-02-18 2017-02-12 
#1.1 2017-02-18 2017-02-13 
#1.2 2017-02-18 2017-02-14 
#1.3 2017-02-18 2017-02-15 
#1.4 2017-02-18 2017-02-16 
#1.5 2017-02-18 2017-02-17 
#1.6 2017-02-18 2017-02-18 
#2 2017-02-25 2017-02-19 
#2.1 2017-02-25 2017-02-20 
#2.2 2017-02-25 2017-02-21 
...

來源

2017-06-19 15:12:57 Sotos

非常感謝您的幫助。 –

library(data.table)

dt <- data.table(V1=as.Date(c("2/18/2017","2/25/2017","3/4/2017","3/11/2017"),format = "%m/%d/%Y"), 
       V2=as.Date(c("2/12/2017","2/19/2017","2/26/2017","3/5/2017"),format = "%m/%d/%Y")) 
for(i in 0:6){ 
    dt[,paste0("colomn_i",i):=V1-i] 
} 
dt[,V2:=NULL] 
temp <- melt(dt,id.vars = "V1") 
setorder(temp,V1,value) 
temp[,variable:=NULL]

即使最終V2，不需要

來源

2017-06-19 15:14:20 quant

下面是使用功能從dplyr和lubridate一個例子。 dt2將是最終的輸出。

# Create example data frame 
dt <- read.table(text = "rptdate st 
2/18/2017 2/12/2017 
2/25/2017 2/19/2017 
3/4/2017 2/26/2017 
3/11/2017 3/5/2017 
3/18/2017 3/12/2017 
3/25/2017 3/19/2017 
4/1/2017 3/26/2017 
4/8/2017 4/2/2017 
4/15/2017 4/9/2017 
4/22/2017 4/16/2017 
4/29/2017 4/23/2017 
5/6/2017 4/30/2017 
5/13/2017 5/7/2017 
5/20/2017 5/14/2017 
5/27/2017 5/21/2017 
6/3/2017 5/28/2017 
6/10/2017 6/4/2017", 
       header = TRUE, stringsAsFactors = FALSE) 

# Load packages 
library(dplyr) 
library(lubridate) 

# Process the data 
dt2 <- dt %>% 
    mutate(rptdate = mdy(rptdate), st = mdy(st)) %>% 
    rowwise() %>% 
    do(data_frame(rptdate = rep(.$rptdate[1], 7), 
       st = seq(.$st[1], .$rptdate[1], by = 1))) %>% 
    mutate(rptdate = format(rptdate, "%m/%d/%Y"), 
     st = format(st, "%m/%d/%Y"))

或者你可以從tidyverse使用map2和unnest功能。

# Load packages 
library(tidyverse) 
library(lubridate) 

# Process the data 
dt2 <- dt %>% 
    mutate(rptdate = mdy(rptdate), st = mdy(st)) %>% 
    mutate(st = map2(st, rptdate, seq, by = 1)) %>% 
    unnest() %>% 
    mutate(rptdate = format(rptdate, "%m/%d/%Y"), 
     st = format(st, "%m/%d/%Y"))

來源

2017-06-19 15:16:44 www

填充日期並在數據框中創建新行

回答

相關問題