on 01-25-2007 4:09 PM
Hello experts,
I am not sure if my approach works so let me elaborate:
ODS1 contains trips on participant level.
The data in ODS1 looks like this:
ODS1:
trip_id___participant_id___price_____trip date
__101_____1___________25.00_____20070112
__101_____27__________12.50_____20070112
__102_____21__________55.00_____20070124
__102_____7___________55.00_____20070124
__102_____15__________35.00_____20070124
Now I want to update the data to InfoCube IC_TP which looks like this:
IC_TP:
dimensions:
- time
- participant_id (has nav. attribute participant_type on which I want to report or do I have to select participant_type instead of participant_id as dimension???)
measures:
- price
- number of trips (according to the data in ODS1 it should be "2" for jannuary)
- number of participants (according to the data in ODS1 it should be "5" for jannuary)
is this updating approach possible?
thanx
hiza
The aproach is posible.
map date and participant_id directly in update rules, key figures as sum and number of participans always 1.
If participant_type is navigational attribute you can do reporting based on it.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
If you use my solution, you should get everything you need. You will need to add the navigational attribute of part_type to the ODS. But you need to make sure that does not change independently of the transactional data. You will also have issues when counting trips, regardless of part_type.
As for the query solution, you can use query 1, to obtain all trip_id's or participants associated to a part_type, then, it will input those trips/parts, into query2 selection to get the number of participants/trips. But, you will run into issues with the part_type because this solution will only identify the number of participants in total, not by part_type. Make sense? I think the query solution with exception aggregation should work. If it doesn't, you may need to create two infocubes, to suffice each solution (number of trips/participants).
Cheers!
/smw
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
That is fine, just create update rules from the ODS (new) to an infocube.
/smw
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Sorry, ie issues.
Message was edited by:
Steve Wilson
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Sorry, IE issues....
Message was edited by:
Steve Wilson
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Create a variable in which you use a query against your first ODS to get the trip_ids and or dates via participant type, then use that variable to select the data from the second ODS. This can all be done at the query level of the second ODS.
/smw
Message was edited by:
Steve Wilson
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Hiza,
this is possible, but you need to add trip_id in your cube if you want to report on it...
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
@Oscar: what do yo mean by "number of participans always 1"? would be nice if you could give me some code example in abap!
@Steve: that was my first approach. Problem is that participants are turned into a measure so I cant report anymore one participant_type!
@Roberto: i don't need to report on trip_id since I have other characteristics like trip_destination etc. in each line of ODS1. I just leaft them out to make the example more simple.
hiza
Hi Oscar,
how can I set 1 as fix value in update rules between ODS1 and Cube in order to count number of participians?
in update rules I have key figures-section, charchteristics-section and time-section:
- For participant_id i choose in charchteristics-section direkt update
- for date I choose also direkt update in time-section
- in key figures-section I choose sum for price and for number of participans I set abab-code in routine "RESULT =1."
But loading data into cube didn'work!!
I can only choose between summation and no updating key figures-section! There is no update mode for setting a constant like in update rules for ODS-Objects!
And how do I count number of trips?
hiza
Hi,
Try this:
For the KF Info object : number of trips,
Take Exception Aggregat = Last Value
Agg.referen.char = Trip_id
And
For the KF Info object : number of participants,
Take Exception Aggregat = Last Value
Agg.referen.char = participant_id
With rgds,
Anil Kumar Sharma .P
Message was edited by:
Anil Kumar Sharma
Hi,
You have take these prperties in the edit screen of the info objects. In the <i>Change</i> screeen , go to <i>Aggregation </i>Tab page. Here you can find these settings.
And try like this .
Hi,
Try this:
For the KF Info object : number of trips,
Take Exception Aggregat = <b><i>Counter(Values unequal to zeros)</i></b>
Agg.referen.char = Trip_id
And
For the KF Info object : number of participants,
Take Exception Aggregat = <b><i>Counter(Values unequal to zeros)</i></b>
Agg.referen.char = participant_id
With rgds,
Anil Kumar Sharma .P
Message was edited by:
Anil Kumar Sharma
With rgds,
Anil Kumar Sharma .P<b></b>
Message was edited by:
Anil Kumar Sharma
My suggestion would be to create another layer of ODS objects.
1 new ODS objects.
(A)One with Trip ID, trip date and price, along with another KF, called Participant Counter. Make the counter and price "additive" in the ODS. For each record, increment the counter. (This gives you number of participants per trip) When you do a number of records, per date, you will get the number of trips. Both KF's would be available for this report.
Detail would need to remain in your current ODS for who the participants were...
trip_id___PartCounter___price_____trip date
__101_____2___________37.50_____20070112
__102_____3__________145.00_____20070124
you could do the same type scenerio with infocubes.
Hope this helps!
/smw
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
81 | |
24 | |
11 | |
9 | |
7 | |
5 | |
5 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.