Transform Stage¶
The transform stage is where data from all of the individual extracted source files is blended together to form a holistic representation of the study using knowledge of the concept associations needed to fully represent entities in your data (e.g. Participants, Biospecimens, that sort of thing).
In this stage you merge the extracted dataframes together into either one large denormalized table (the “default” output) or into smaller subset tables where each row in a subset table fully represents one entity.
Note
It should be possible to one day fully automate this process according to formal rules about how entities relate to each other in your target service, but for now that is left as an exercise for the reader.