Hi,
Thanks for your suggestions.It is really helpful to understand concepts for person from non BI background.
Below are my observations.
a.) Final node with semantics will come default with aggregation since as per performance t will be better if we aggregate data if we are using cube else it will act as a general node if we are not giving any measures and don't have filter option.
b.) Diff between aggregation vs projection is that projection will have only apply filter option and dont have '.output as aggregated column' and latter have the same and if we don't want to aggregate we can go for Projection view only.
I have one more query. I worked on one sales report in which we need to join multiple sales table and to filter the data. I used calculation view and then used projection node and applied filters and on top of that applied joins. But after multiple projections and joins it gave me error ' Out of memory'.
Then I did the same in attribute view by simply added the tables in data foundation and applied joins and filters and data retrieved very quickly.
I don't know why same not happened with Calculation view why its throwing 'Out of memory' error?
BR
Sumeet