Dates in Excel's internal format

[pitakakariki]

I have a column which has dates in Excel's internal format. The Excel file comes from someone's reporting software, so I can't just format the column in Excel (unless I want to do it every time I get an update).

◢	Date
1	39448
2	39449
3	39450

date_test.xlsx

These should map to the first, second, and third of January 2008.

> as.Date(read_excel("../notes/date_test.xlsx")$Date, origin="1899-12-30")
[1] "2008-01-01" "2008-01-02" "2008-01-03"

But if I tell read_excel that it's a date:

 z <- read_excel("../notes/date_test.xlsx", col_types="date")

Warning messages:
1: In read_xlsx_(path, sheet, col_names = col_names, col_types = col_types,  :
  [2, 1]: expecting date: got '39448'
2: In read_xlsx_(path, sheet, col_names = col_names, col_types = col_types,  :
  [3, 1]: expecting date: got '39449'
3: In read_xlsx_(path, sheet, col_names = col_names, col_types = col_types,  :
  [4, 1]: expecting date: got '39450'

> z
  Date
1 <NA>
2 <NA>
3 <NA>

[jennybc]

I have a column which has dates in Excel's internal format.

So this isn't quite true. Generally, a date cell in xlsx will carry a date style, which these cells do not. They are just integers and it is a choice to interpret them as dates. But I suppose we can interpret numbers as dates, when explicitly directed to do so via col_types = "date".