Proper use of GROUP BY

Question

I wrote the view below to query orders that have shipped quantity from selected warehouse subinventories. I need to compute the total of unique orders and sum for a given date along with extended price.
I have tried many SELECTs from SCO_REQSGROSS_V including
select TRXN_DATE
, count(*)
, SUM(EXT_PRC)
from SCO_REQSGROSS_V
where trxn_date = to_date('30-AUG-2010')
group by TRXN_DATE, order_num
I need a count of all the unique order numbers and there ext_prc for each date in a given date range. I keep getting back the line counts. There is something wrong with my approach using the GROUP BY clause. Do I need to use a partition clause and how would this written?
create or replace view SCO_REQSGROSS_V (
TRXN_DATE
, order_num
, EXT_PRC)
AS
select TRUNC(transaction_date)
, segment1
, ext_price
from
-- lines
SELECT DISTINCT prh.segment1
, prl.line_num
, ool.ordered_item item_number
, mmt.trx_source_line_id
, mmt.transaction_reference
, ooh.order_number
, prl.quantity
, ROUND(ool.shipped_quantity * prl.unit_price,2) ext_price
, mmt.transaction_date
, mmt.subinventory_code
FROM apps.po_requisition_headers_all prh
, apps.po_requisition_lines_all prl
, apps.oe_order_headers_all ooh
, apps.oe_order_lines_all ool
, apps.mtl_material_transactions mmt
WHERE prh.requisition_header_id = prl.requisition_header_id
AND ool.header_id = ooh.header_id
AND prh.requisition_header_id = prl.requisition_header_id
AND ooh.source_document_id = prh.requisition_header_id
AND ool.source_document_line_id = prl.requisition_line_id ----(+)
AND ool.line_id = mmt.trx_source_line_id
AND mmt.transaction_type_id = 53 -- Internal Order Pick
AND prl.source_type_code = 'INVENTORY'
AND prh.authorization_status = 'APPROVED'
AND prl.item_id = ool.inventory_item_id
AND ool.shipped_quantity IS NOT NULL
AND ool.subinventory = mmt.subinventory_code
AND mmt.subinventory_code IN ('APPLIANCE', 'ELECTRICAL', 'ELEVATOR', 'FURNITURE', 'HARDWARE', 'HEAT'
, 'JANITORIAL', 'JANIT 1', 'OFFICE', 'PAINT', 'PAPER RM', 'PLUMBING', 'SKIL', 'SURPLUS', 'TOOL', 'UNIFORM')
)

Answer 1

I ran this version of your code:
select distinct trxn_date
count(order_num) over (partition by trunc(trxn_date)) cnt_orders,
sum(ext_prc) over (partition by trunc(trxn_date)) total_per_txn_day -- for total_prc per day
from SCO_REQSGROSS_V
where trunc(trxn_date) = to_date('30-AUG-2010');
it returned one row and total_per_txn_day was correct at 51044.21. However cnt_orders was 303. This is nominally correct but not really what I am after because what I need is a count of unique order numbers.
The logic
SELECT * FROM
FROM SCO_REQSGROSS_V
where trunc(trxn_date) = to_date('30-AUG-2010')
order by order_num;
returns some duplicate order numbers. I want to count the unique order numbers but still include the ext_prc for all of the orders even the duplicates.
Here is the first several rows of output from my query above. Desired cnt_orders is 3 and the total_per_txn_day is 1528.98:
Date..........Order#.....Ext_Prc
30-AUG-10     399884     90
30-AUG-10     399893     108
30-AUG-10     403664     114.48
30-AUG-10     403664     68.7
30-AUG-10     403664     100.02
30-AUG-10     403664     114.48
30-AUG-10     403664     139.04
30-AUG-10     403664     43.04
30-AUG-10     403664     49.8
30-AUG-10     403664     120
30-AUG-10     403664     55
30-AUG-10     403664     198.41
30-AUG-10     403664     129.6
30-AUG-10     403664     198.41
Thank you.