New generation technique requires datatypes to be explicitly specified so that the correct estimator can be chosen. At the moment, data must be passed in to do this, but we could easily specify the datatype of each node as metadata in the DOT file.
It would also be good to specify hidden variables as part of the DAG too.
New generation technique requires datatypes to be explicitly specified so that the correct estimator can be chosen. At the moment, data must be passed in to do this, but we could easily specify the datatype of each node as metadata in the DOT file.
It would also be good to specify hidden variables as part of the DAG too.