ScalaSparkCompute does not work when input schema is a macro. Validation also fails, however you can workaround the validation by setting the code as a macro as well - but pipeline run still fail. Both scenarios should be fixed.
Schema to set in the runtime arguments:
Scala code: