I have timestamps in a column which I have imported in SPSS. Example, 7/6/2011 2:21 in a column called 'Observation'
This is in the string format. Now I also have timezone corrections for these data. So, -60 would mean subtract 60 minutes from this date.
How would I do this in SPSS syntax?
There are native date formats in SPSS, but unfortunately it does not appear that any cover the example you posted. I would parse the beginning of the string field to get the mm/dd/yyyy and the hh:mm part seperate, convert those into their representative time formats, and then do the time calculations.
For an example
data list fixed / observation (A25).
begin data
7/6/2011 2:21
10/11/2011 15:42
07/06/2011 02:21
3/15/2011 0:21
end data.
*getting the data part, assuming the space will always delimit the two parts.
compute #space = char.index(observation," ").
string date (A10).
compute date = char.substr(observation,1,#space-1).
*getting the time part.
string time (A5).
compute time = char.substr(observation,#space+1,5).
execute.
*now converting them into date formats.
alter type date (A10 = ADATE10).
alter type time (A5 = TIME5).
*you should check these carefully to make sure they were converted correctly.
*now making one time variable.
compute date_time = date + time.
formats date_time (DATETIME17).
execute.
*now it is just as simple as subtracting the specified value.
compute date_time_adj = DATESUM(date_time,-60,"minutes").
execute.
formats date_time_adj (DATETIME17).