Since we can select features from different feature groups to create our own training datasets. I wonder how to join those selected columns?is it by a common primary key? what’s the principle behind it?
If you don’t specify anything when joining Hopsworks will use the largest matching subset of primary keys with the same name. You can also specify on which key you would like to join, thus overwrite the joining key.
You will find more details in this notebook: https://github.com/logicalclocks/hops-examples/blob/master/notebooks/featurestore/hsfs/feature_exploration.ipynb