CODEWITHSUNDEEP
rdplyr
EML
EML

Reputation: 671

Converting numeric column to date and encountering: "Error in as.Date.numeric(value) : 'origin' must be supplied"

I am attempting to convert several columns to dates if they follow the format of "days since origin" (here, if nchar(col)==5) and encountering an error:

Example data:

df <- data.frame(col1=c(44008, 440017, NA, 77), col2=c(43476, 43479, 77, NA))
> df
    col1  col2
1  44008 43476
2 440017 43479
3     NA    77
4     77    NA

Attempt:

df %>%
mutate_at(vars(col1, col2), funs(as.numeric(as.character(.)))) %>%
  mutate_at(vars(col1, col2), 
            funs(case_when((nchar(.)==5)~(as.Date(., origin = "1899-12-30")),TRUE~.)) )

Error message:

Error in as.Date.numeric(value) : 'origin' must be supplied

Desired:

        col1       col2
1 2020-06-26 2019-01-11
2 3104-09-21 2019-01-14
3       <NA>         77
4         77       <NA>

Why am I getting an error if I provided the origin date? Thank you in advance.

Upvotes: 1

Views: 96

Answers (1)

akrun
akrun

Reputation: 887691

If the intention is to convert only those elements with number of characters at least 5 to Date class and keep the others as such. But, in the process, we cannot have a Date class as numbers such as 77 can be either numeric or string. Here, we convert to string with as.character

library(dplyr) # v >= 1.0.0
df %>% 
   mutate(across(everything(),
    ~ case_when(nchar(.) >= 5 
           ~ as.character(as.Date(., origin = '1899-12-30')),
       TRUE ~ as.character(.))))
#     col1       col2
#1 2020-06-26 2019-01-11
#2 3104-09-21 2019-01-14
#3       <NA>         77
#4         77       <NA>

Upvotes: 1

Related Questions