I agree, very odd. 40030 is actually that SAS date for 08/06/2069 which ia a little too similar to 08/05/2009 for me to think it is just coincidence. Is there anything else you can offer to help explain this?
--------------------------------------------------
data _null_;
a = 40030;
format a YYMMDD10.;
put a;
run;
-------------------------------
2069-08-06
well..I'm just using the folliwng code:and it results in the date conversion.
I even tried the proc import and it resulted the same.
%macro test;
proc sql;
connect to excel (path="&inpt.\Walgreens 2009 08.xls"
header=no mixed=yes );
create table walgreens as
select * from connection to excel
(select * from [%sysfunc(dequote(%superQ(wksht)))]);
disconnect from excel;
quit;
%mend test;
%test;
I would check the Excel table to make sure that all of the dates in that column are entered in the same manner. If they have different formats, that might explain this. If not, I think I would give this one to tech support.
sounds like SAS is reading the internal exccel value, without recognising the datatype and .applying the standard adjustment (1/1/1960-1/1/1900 with the fraction representing the fraction of the day at the time point.
Although we can fix it, I hope SAS Customer Support can provide the option to enable appropriate conversion to be automatic