Graphing of Timestamps from PostgreSQL and external program - excel

Question

I am looking for a way to graph the time a Toy Store stock boy was checking into his job. There would be a defined start and end time for the overall graph but looking to span the amount of time he/she spent on the job.
The actual database would simply have the times the stock boy checked into work and the time he/she clocked out. Example:
timeshifts table
employerID | start_time | end_time
---------------------------------------------
1 | 2014-12-10 09:00:00 | 2014-12-10 09:37:00
1 | 2014-12-10 09:53:00 | 2014-12-10 11:44:00
1 | 2014-12-10 12:00:00 | 2014-12-10 15:00:00
after extracting the data and importing into (??), my IDEAL graph output would look something like
I know PSGL can't do this on its own but not sure if I need to structure the data in any special format to calculate the X distance (something like an age(end_time,start_time) or the like).
Thank you for your help ahead of time!

Answer 1

In PostgreSQL, you can subtract timestamps.
select employer_id, start_time, end_time, end_time - start_time as elapsed_time
from timeshifts;
employer_id start_time end_time elapsed_time
--
1 2014-12-10 09:00:00-05 2014-12-10 09:37:00-05 00:37:00
1 2014-12-10 09:53:00-05 2014-12-10 11:44:00-05 01:51:00
1 2014-12-10 12:00:00-05 2014-12-10 15:00:00-05 03:00:00
Whether Excel can recognize the values in "elapsed_time" is a different thing. It might be easier to do the subtraction in Excel.
create temp table timeshifts (
employer_id integer not null,
start_time timestamp with time zone not null,
end_time timestamp with time zone not null,
check (start_time < end_time),
primary key (employer_id, start_time)
);
insert into timeshifts values
(1, '2014-12-10 09:00:00', '2014-12-10 09:37:00'),
(1, '2014-12-10 09:53:00', '2014-12-10 11:44:00'),
(1, '2014-12-10 12:00:00', '2014-12-10 15:00:00');