Following on from Nash Gajic’s excellent blog on connecting HCP to Cloud for Analytics I wanted to get this working for myself with some of my own analytical datasets.
Before connecting HCP into C4A and getting started on building a story I spent a little bit of time getting some tables, data and HANA analytic models (attribute and analytic views) setup in the HCP database to give me some data sets to work with. I had a couple of data set to work with, the good old eFashion (the classic BusinessObjects demo based on retail data) and EIM360 (SAP ERP profitability analysis data). I had some table/data exports (binary files) from an on premise HANA system that I was able to use.
With these binary exports to hand, and getting my HCP system connected HANA studio, I was able to import the files into my HCP instance to get the tables and data uploaded. I then created some Attribute, Analytic & Calculation views on top of the tables.
I decided to use the HANA studio application on my laptop rather than the HCP web IDE to build the views as from what I could see I could only create HANA calculation views in the web IDE rather than Attribute or Analytic Views (at least not graphically anyway). I did double check to see everything was there in the HCP web IDE once I had done the activities in the studio (below is a list of the view in the web IDE).
I then created a remote system connection in C4A to my HCP system and then created a C4A model by connecting to the HANA Analytic View I created on top of the eFashion tables & data. I was able to then to go and build a story on my new C4A model. The process of connecting HCP and creating the C4A model was extremely easy.
The eFashion & EIM datasets can be accessed vie the URLs below if others want to use them (download, unpack and choose Import from within the Catalog/Schema area within the HANA Studio). You can then build the HANA views on top. The eFashion views are probably the easiest. I created 3 Attribute Views for Article, Store and Calendar (each using the relevant base table) and then an Analytical view using the fact table and then linking the 3 Attribute views.