How to create Hierarchy Columns

User_DX76J

Need Guidance on How to create hierarchy columns in Oracle Analytics Server

Goal:

Currently, I am trying to make graph that can drill up/down the values. For instance, if I drill to attribute in x-axis from year to quarter to month, and so on.

Current Progress:

I tried to import dataset from excel or csv (or even create a dataset from Oracle Database Connection). It does not show the hierarchy columns of date as I expected. The result is in the below picture. With this, I cannot drill up/down columns as I want in graph.

Reference source(s):

I find source from this YouTube link, it already have the hierarchy columns of date as I intended. However, it does not show the steps on how to create the hierarchy columns.

https://youtu.be/PBxADXChCqw

My questions are:

1. How to create hierarchy columns of date in Oracle Analytics Server? (or does it has any other ways that I can drill up/down the columns of date in graph without using hierarchy columns?)
2. Do I need to use Subject Area to create hierarchy columns?, What is the importance and difference between normal dataset and using Subject Area?

Federico Venturin

Hi @User_DX76J ,

You can create hierarchies in the RPD/semantic model using the Model Administration tool or the Semantic Modeler. Refer to this page for step-by-step instructions (it's old, but still valid): https://www.oracle.com/webfolder/technetwork/tutorials/obe/fmw/bi/bi1221/rpd/rpd.html

In DV, you can also right-click on a given attribute column in visualisations, and select the Drill To Attribute/Hierarchy option. Please note that this option allows you to drill to any other attribute in the dataset/subject area.

The goal of the RPD/semantic should be to provide a unique source of truth for data accross the organization and provides several enterprise level features such as horizontal/vertical federation, row-level security, localization options, etc. On the other hand, datasets are easier to create/manage for new users, include a powerful editor to transform and enrich data, but do not offer the same leve of enterprise features as the RPD/semantic model.