添加日期到明年 1 月的行

Question

我确实有相当复杂的案例需要我解决。让我通过一个例子来解释你..所以我们从下表开始：

 Datum Urlaub_geplannt
1 2018-10            1410
2 2018-11             940
3 2018-12             470


structure(list(Datum = structure(1:3, .Label = c("2018-10", "2018-11", 
"2018-12"), class = "factor"), Urlaub_geplannt = c(1410, 940, 
470)), .Names = c("Datum", "Urlaub_geplannt"), row.names = c(NA, 
-3L), class = "data.frame")

我希望在明年 1 月之前将新行添加到此表中（基准列），并且所有其他列都应填充为 0。在这种情况下，最终表应如下所示：

 Datum Urlaub_geplannt
1 2018-10            1410
2 2018-11             940
3 2018-12             470
4 2019-01             0

但是，随着我的数据发生变化（实际上是在Shiny 中），以某种方式自动使其成为“年末”非常重要。

我的意思是，如果我有来自 2019 年的行的新数据，我希望自动获得“结束日期”为 2020 年 1 月。感谢您的帮助！

Answer 1

基本 R 方法

get_date_till_Jan <- function(df) {
  #Convert the character dates to actual Date objects
  max_Date <- max(as.Date(paste0(df$Datum, "-01")))

  #Get the date for next year January
  next_Jan <- as.Date(paste0(as.numeric(format(max_Date, "%Y")) + 1, "-01-01"))

  #Create a monthly sequence from the max date to next Jan date
  new_date <- format(seq(max_Date, next_Jan, by = "month")[-1], "%Y-%m")

  #Create a new dataframe with all values as 0 and change only the Datum 
  #column with new_date and rbind it to original dataframe
  rbind(df, transform(data.frame(matrix(0, nrow = length(new_date), 
      ncol = ncol(df), dimnames = list(NULL, names(df)))), 
      Datum = new_date))
}

df <- get_date_till_Jan(df)
df

#    Datum Urlaub_geplannt
#1 2018-10            1410
#2 2018-11             940
#3 2018-12             470
#4 2019-01               0

这适用于任意数量的列

df['another_col'] = 1:4
get_date_till_Jan(df)


#     Datum Urlaub_geplannt another_col
#1  2018-10            1410           1
#2  2018-11             940           2
#3  2018-12             470           3
#4  2019-01               0           4
#5  2019-02               0           0
#6  2019-03               0           0
#7  2019-04               0           0
#8  2019-05               0           0
#9  2019-06               0           0
#10 2019-07               0           0
#11 2019-08               0           0
#12 2019-09               0           0
#13 2019-10               0           0
#14 2019-11               0           0
#15 2019-12               0           0
#16 2020-01               0           0

Answer 2

dplyr 和 full_join 的解决方案：

library(dplyr)
library(lubridate) # for ymd() function


d <- d %>% 
  mutate(Datum = paste0(Datum,"-01"),
         Datum = ymd(Datum)) # correct Date format

min_year <- year(min(d$Datum))
min_date <- min(d$Datum)

# create a data.frame of possible dates
fill_dates <- data.frame(Datum = seq.Date(
  min_date, # min date avaiable
  as.Date(paste0(min_year+1,"-01-01")), # until first Jan next year
  by = "month"))

现在我们可以加入两个data.frames：

d %>% 
  full_join(fill_dates, by="Datum") %>% # full_join of the two tables
  # the full_join will add all new row not present in d originally, with NA
  mutate(Urlaub_geplannt = ifelse(is.na(Urlaub_geplannt), 0, Urlaub_geplannt))

#       Datum Urlaub_geplannt
# 1 2018-10-01            1410
# 2 2018-11-01             940
# 3 2018-12-01             470
# 4 2019-01-01               0

数据：

d <- structure(list(Datum = structure(c("2018-10", "2018-11", 
                                                      "2018-12"), class = "character"), Urlaub_geplannt = c(1410, 940, 
                                                                                                         470)), .Names = c("Datum", "Urlaub_geplannt"), row.names = c(NA, 
                                                                                                                                                                      -3L), class = "data.frame")

Answer 3

df <- structure(list(Datum = structure(1:3, .Label = c("2018-10", "2018-11", 
                                                       "2018-12"), class = "factor"), Urlaub_geplannt = c(1410, 940, 
                                                                                                          470)), .Names = c("Datum", "Urlaub_geplannt"), row.names = c(NA, 
                                                                                                                                                                       -3L), class = "data.frame")




Datum <- format(seq.Date(as.Date(paste0(df$Datum[nrow(df)],"-01")),
                         as.Date(paste0(substring(seq.Date(as.Date(paste0(as.character(df$Datum[1]),"-01")), 
                                                           length = 2,
                                                           by = 'year')[2],1,4),"-01-01")),
                         by = "month"

),"%Y-%m")


new_df <- data.frame(Datum  = Datum, Urlaub_geplannt = rep(0,length(Datum)))


total_df <- rbind(df,new_df)

total_df
#>     Datum Urlaub_geplannt
#> 1 2018-10            1410
#> 2 2018-11             940
#> 3 2018-12             470
#> 4 2018-12               0
#> 5 2019-01               0