Combine datasets into recipes . Add data transformations.
Recipes always start with a dataset or more and result in a new dataset.
Create Recipe from dataset you are starting with.
For custom ordering of records, you can bucket a field.
Bucketing Dates
"Append data" feature, appends rows. For matching field names from two sources, it puts data and for non matching fields it puts empty.
"Add Data" feature adds column to existing dataset from another dataset. It will work on matching lookup keys
Create a formula to perform operation on two columns
After these transformations, click on "Create Dataset" button.
Keep only the required fields
Recommend to add related fields/objects while creating a dataset.
Start always with lower grain (example opportunity in opp==>Account relationship)
Design base level granular datasets and than combine them.
Inner Join: Example Opportunity Inner Join with Cases. Duplicates opportunity and case records in below table as account1 have two opportunity and two cases.
Left Join : Example Opportunity Left Join with Cases. Inner join + Not matching entries in left table.
Right Join : Inner join + Not matching entries in right table.
Full Outer Join : Inner Join + Not matching entries in left table + Not matching entries in right table
Single Lookup : Just one match + Not matching entries in left table
Left, Right, Full, Single Lookup - None of these solved problem of opportunity and cases.
Lower Grain Left Join with higher grain makes sense.
Bar - GroupBy , Cluster.
