I have a number of customized Python View Nodes which I use regularly. Till now I stored all Nodes into a “dummy workflow” and copy them if needed.
Is there a more clever way like store these nodes in the Node Repository directly? Google said there is a “Custom Node Repository” but only as a server extension. (I use Knime 4.1.4)
you are right, Custom Node Repository is a feature of KNIME Server. However if I got what you are looking for and what Custom Node Repository is offering it’s not a match. You want to save a node with a configuration and then every time you need it drag into a workflow while in Custom Node Repository gives you possibility to only customize repository without nodes being already configured.
Besides you current workaround you could create Component for each customized node you want to have and share it.
Additionally there is feature request for default node configuration (internal reference: AP-11674) and will add your request to it.
Could the python nodes maybe be in a local workflow which gets called? Not sure about that.
hmm… Not sure. Maybe but that is still a hacky workaround.
thanks. I am not sure what the benefit of a component is, because my Python Node is one node only and not a bunch of instructions.
The benefit of the component is that you can share it. You can make a folder in your workspace and name it “Components” and then when sharing a component save it into that folder. Then when needing to use it, you can pull it into your workflow.
On top of that if you make a change to a component upon opening the workflows using it, you may update them. eg. 1 place of truth.
doesn’t matter if it’s only one node. It is important you get needed functionality from it.
Adding two links for more on Components, documentation and blog post on it:
This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.