I am reading Excel Workbooks (.xlsx format) with read.xlsx() from the xlsx package. My Workbooks have dates stored as functions (e.g., =DATE(1900, 1, 1)) rather than strings or fractional days from 1899-12-31 (e.g., 1/1/1900 or 1).
This is not a problem for read.xlsx if the Workbook has ever been opened/evaluated. But many of these Workbooks have never been opened/evaluated (they're generated by my data provider, not by Excel). Because the Workbook has not been opened/evaluated it contains -1 rather than the actual fractional number of days since 1899-12-31 (e.g., 37036 for May 25, 2001). I learned this here. So if I read a date in one of these (never opened) Workbooks, I get a column of -1, which read.xlsx interprets as 1899-12-30 (i.e., one day before 1899-12-31).
Is there a fix here? Can I touch a Workbook to force evaluation? Is there a bit that I can flip in read.xlsx to force evaluation first? I see only a colClasses argument, which doesn't seem to be the problem? Can I avoid using an Excel VBA script to write these Workbooks to text? Sorry I can't provide any reproducible code, but if you open an Excel Workbook to enter =Date(), the problem won't exist.
Using XLConnect package , You can use :
getForceFormulaRecalculation(object,sheet)
to compute for all sheets :
getForceFormulaRecalculation(wb, sheet = "*")