possible to support sec level of granularity?

[hokiegeek2]

Hi, was wondering if it is possible update this algorithm for seconds level of granularity?

[Mário Teixeira]

Hey @hokiegeek2

In my case, when wanting to evaluate two levels of granularity, in fact we still only require a single analysis to be made.
An example to be clearer:

Having a dataset of
Dimension_1 | Dimension_2 | KPI
X1 | Y1 |10
...

What we want to understand is if on combination X1+Y1 there is an outlier. Therefore transforming the dataframe to only have:
Dimensions | KPI
X1,Y1 | 10

performs the exact same intent of the analysis, while passing only a single level of granularity.

Sorry for typos or formatting of the answer, hope it was helpful :)