R数组的日期转换

Question

I get an array of dates from a database that I would convert to Date types.

> dates
              start            finish
1 29-Oct-2017 00:00 30-Oct-2017 00:00
2 30-Oct-2017 00:00 31-Oct-2017 00:00
3 31-Oct-2017 00:00 01-Nov-2017 00:00
4 01-Nov-2017 00:00 02-Nov-2017 00:00
5 02-Nov-2017 00:00 03-Nov-2017 00:00
6 03-Nov-2017 00:00 04-Nov-2017 00:00

I can convert each column dates['start'] and dates['finish'] separately with

> as.Date(dates['start'][,1], format="%d-%b-%Y %H:%M")
[1] "2017-10-29" "2017-10-30" "2017-10-31" "2017-11-01" "2017-11-02"
[6] "2017-11-03"

But is there a way to do it in one step?

I tried both rows simultaneously, bu that doesn't work.

> as.Date(dates[c('start', 'finish')], format="%d-%b-%Y %H:%M")
Error in as.Date.default(dates[c("start", "finish")], format = "%d-%b-%Y %H:%M") :
  do not know how to convert 'dates[c("start", "finish")]' to class “Date”

Answer 1

According to ?as.Date

The as.Date methods accept character strings, factors, logical NA and objects of classes "POSIXlt" and "POSIXct".

Assuming that the dataset is data.frame , it would not work as the requirements are not met. We can use lapply to loop over the columns, apply the as.Date to convert to Date class and assign the output back to the columns of interest

dates[c('start', 'finish')] <- lapply(dates[c('start', 'finish')], 
         as.Date, format = "%d-%b-%Y %H:%M")

---

With tidyverse , we can use mutate_at

library(dplyr)
library(lubridate)
dates %>%
      mutate_at(vars(start, finish), dmy_hm)