How to set measure values for dimension levels

Hi team,

We are trying to import some REST data into a separate eazyBI cube (so, right now, not linked with Issues cube for the purpose of this question), but we are having an aggregation issue I will try to explain.

I will put a simple example to show the problem, though the data we have is much bigger:

domain product component myMeasure
domain1 productA componentX 1
domain1 productA componentY 0

In our model we have 1 single dimension with 3 levels, being domain top level, then product and finally component. EazyBI default aggregation formula is sum. Thus, importing this data results in productA (and domain1) having a myMeasure of 1. The problem we have here is that our data contains dozens of measures, whose correct aggregation is not always meant to be a sum; we have measures that aggregate with sum, avg, weighted avg and other complex formulas. Let’s say then for the sake of this example we want myMeasure to aggregate using avg formula, so its value for productA (and domain1) should be 0.5.

We could rely on creating custom calculated measures to perform this calculus. But, if possible, we would want to achieve this by directly importing our desired values into the upper levels (domain and product). REST api will provide these. So, following the example, this is the data we want to import:

domain product component measure
domain1 productA componentX 1
domain1 productA componentY 0
domain1 productA 0.5
domain1 0.5

Sadly, importing this data is not showing us the results we would want on eazyBI.

Here, the same data inside the data import on eazyBI:

And here the results on a report:
image

We can understand somehow the reason why we have those values (we are importing a separate productA/domain1 line with a value that gets summed up with the default aggregation sum value), but we fail to see a solution to fix it. We experimented a bit with the advanced options but no luck so far.

Is there any way to achieve this? Is there any option to “choose” aggregation value at import?

Thanks,
Guillermo

Hi @guillermo.esteban,

Unfortunately, eazyBI doesn’t offer other aggregation methods in data mapping. Thus, the averages have to be calculated using calculated measures - for example, dividing the measured value by the “Component” item row count.

For more complex metrics, each “Component” value needs to be considered. This is partly due to the possibility of filtering the reports by other dimensions, which would then provide wrong results if the average is already precalculated on the top “Domain” level.

Best,
Roberts // support@eazybi.com

1 Like

Hi Roberts,

I was not taking into account the filtering issue you talk about. What you say makes a lot of sense and I can see why this is not supported.

Thanks for your response.

1 Like