conditional conversion from character to date for a dataframe column in r

问题

I have a dataframe I read from an excel file, like below. Date turned out to be in 5-digits format or a date string format.

df = data.frame(Date = c('42195', '3/31/2016', '42198'), Value = c(123, 445, 222))

Date     Value
42195      123          
3/31/2016  445          
42198      222  

I want to clean up the column and convert everything into date format. I did the following.

df %>% 
  mutate(Date = ifelse(length(Date)==5,as.Date(Date, origin = '1899-12-30'), as.Date(Date) ))

I got error like this:

Error in charToDate(x) : character string is not in a standard unambiguous format

What did I do wrong? I could not figure out why after many attempts to fix it. Thanks a lot.

回答1:

There are few changes that you need to do :

1) Instead of length(Date)==5, I guess you are looking for nchar

2) To change excel date to R date, they need to be numeric. Currently they are factors.

3) You need to provide specific format to as.Date when date is not in standard format.

4) Use if_else instead of ifelse since the latter would change them to numbers.

library(dplyr)

df %>% 
  mutate(Date = as.character(Date),
         Date = if_else(nchar(Date) ==5, 
                   as.Date(as.numeric(Date), origin = '1899-12-30'), 
                   as.Date(Date, "%m/%d/%Y")))

#        Date Value
#1 2015-07-10   123
#2 2016-03-31   445
#3 2015-07-13   222

来源：https://stackoverflow.com/questions/60254796/conditional-conversion-from-character-to-date-for-a-dataframe-column-in-r