Uploading Revised Data

[kiamkhai]

Dear TM1 developers

I would like to seek assitance or views on the following:

1. If I received revised data from the data providers, I will upload into the data into the respective cubes in TM1. My question is who can I differentiate or identify which element has the data been replaced by the revised data given by the data provider from those elements where their data remains unchanged.

2. If the data providers provide the new revised data to us, those elements with revision will be included in the revised version with value or figure and for elements without revision, their data values will be blank instead of "0" or "-". When I upload the revised data into TM1, the elements with revised data will be replaced with the revised figures but those without revision will have the old value or figure becuase they have figures/data previously. By right, this group of elements shouldn't have any figures after the revision. How can I resolve this issue.

Thanks.

Regards
Tee Kiam Khai

[jim wood]

Hello there,

The standard way we handle situations like this in TM1 is clearing the cube first before uploading the new information. Normally I do this by having a TI extract based on a view that export the data to be cleared to a flat file, then another to clear the cube based on the file. Then I load revised information.

This approach gives you:

1) The ability to go back to the original information if needed
2) the ability to control the time span cleared using the view

I hope this helps,

Jim.

[kiamkhai]

Dear Jim

Thanks for your suggestion. In fact we have been doing that alternate too. Is there a better way of adding some scripts in order to refresh revised data instead of updating the revised data.

2. Any suggestion to my second query/problem?

Thanks

Cheer

[Marcus Scherer]

kiam,
to your second problem: you could introduce an additional dimension with elements 'original', 'revised1' .. and a C-Element that calculates the delta through consolidation factor. In reports you would probably present the C-Element but users could always track the revision history .

Another possibility would be to replace existing values directly during upload with TI and skip the records with amount = 0.
Do a CellGetN() first to read the existing cell value and add a CellPutN() for correction. With this approach you would loose history status.

Just some ideas.

Success,

Marcus