The H2O Local Context Node starts a single node H2O instance and provides the connection for KNIME. In order to leverage the highspeed algorithms implemented in H2O, the first step is always importing your data to your H2O instance where it is stored in-memory in H2Os so called "Frames". This is done by the "H2O Table to Frame" Node. The Node expects a data input as well as an H2O Context, provided here by the "H2O Local Context" Node. To work further with the data in KNIME it is necessary to re-import it from H2O to a KNIME Table which can be achieved by using the "H2O Frame to Table" Node. From there you are able to connect the data flow to regular KNIME Nodes.
This is a companion discussion topic for the original entry at https://kni.me/w/2Jx0srnqZzPExgOJ