Polling Datawarehouse Design

Question

Post reply

Polling Datawarehouse Design

Lidou123

Hall of Fame

Points: 3186
More actions
January 26, 2019 at 11:23 am

#379126

Hello
I really need your help.
I am currently working on a BI project for a polling institute.
We have surveys that we do all over the world and I have to create a data warehouse that collects all these polls and analyzes them.
The difficulty I encounter in modeling my datawarehouse is that I have a hundred measures.
I've heard of a modeling in which we put measures as a dimension and we have measures like QTD, YTD.
Would anyone have an example of this type of modeling ???

Viewing 5 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic. Login to reply

frederico_fonseca SSCoach Points: 16157 More actions · Answer 1

possibly not the answer you wish for but what have you researched so far?

100 measures is not much - one of my "cubes" has 22 dimensions, 1 fact and over 350 measures.

Lidou123 Hall of Fame Points: 3186 More actions · Answer 2

frederico_fonseca - Saturday, January 26, 2019 11:47 AM
possibly not the answer you wish for but what have you researched so far?
100 measures is not much - one of my "cubes" has 22 dimensions, 1 fact and over 350 measures.

I would like to have an example of a Datawarehouse design in which the measures are in a dimension table.

Martin Schoombee SSCoach Points: 19128 More actions · Answer 3

Lidou123 - Saturday, January 26, 2019 11:23 AM
Hello
I really need your help.
I am currently working on a BI project for a polling institute.
We have surveys that we do all over the world and I have to create a data warehouse that collects all these polls and analyzes them.
The difficulty I encounter in modeling my datawarehouse is that I have a hundred measures.
I've heard of a modeling in which we put measures as a dimension and we have measures like QTD, YTD.
Would anyone have an example of this type of modeling ???

I've attempted to do this before, but I wouldn't recommend it unless you have no semi-additive or non-additive measures (which you are likely to have at some point). At first it seems to be a good idea to calculate everything in advance, but permutations become a nightmare.

I'd recommend that you rather use Analysis Services cubes for something like this. It is designed to handle 100 measures quite easily. I have a Standard Edition multi-dimensional cube with about 500 measures (most of them calculated measures) and it works pretty well.

Luis Cazares SSC Guru Points: 183706 More actions · Answer 4

I don't see how QTD or YTD are measures. They certainly look as calculations done on measures using a time dimension.

Luis C.
General Disclaimer:
Are you seriously taking the advice and code from someone from the internet without testing it? Do you at least understand it? Or can it easily kill your server?

How to post data/code on a forum to get the best help: Option 1 / Option 2