Word Vector Apply on a document set different from the document set the model was trained on aborts execution with the following error:
ERROR Word Vector Apply 0:309 Execute failed: No tokens in row Row318 match the vocabulary!
I would expect, the node just creates cells with missing values or '0's (at least this is what I see in the source code). Instead it completely aborts execution and prints out the above error message. Is this an intended behavior or a bug?
Yes please make that configurable. I've been struggling with that for 2 days. It will save me from a lot of headaches for my publishing. Could you make it asap?
I'm sorry for the inconvenience. In the upcoming release of KNIME the Word Vector Apply Node will output missing cells if no words match the vocabulary.