Knime 3.2 Slow Puncuation Extractor and N Chars Filters

New to Knime.

Just using the Puncuation Erasure and N Chars Filters for the first time.  However 1600 5-50 line documents are taking over 10 minutes to do the Punctuation Erasure and N Chars Filter.  These are the new (non deprecated) nodes.  Upped the Heap size from 1024m to 2048m in knime.ini.

Working on a 4 core 8GB Ram Apple Mac Book Air.

Thoughts?  Is this a normal pricessing time?  Is this a 3.2 Bug?  Do I need to try some sort of change in my configuration?


In reviewing the work flows on my disk.  I've discovered that the folders on my disk that represent the Puncuation Erasure and N Chars Filter have grown to 2.2 GB each.  I've not seen this with other nodes.  I'm using the new 3.2 Code.  I'm wondering if some sort of bug or debug mode was left on as part of the upgrade?

If I use the nodes in the Extended NER Preprocessing "Meta Nodes" I find that processing time is cut to something more reasonable.  My guess is that this is some sort of bug...  Can someone clarify the nature of this issue?  And how resolution of Node problems is handled in this community?  Thanks...

Hello Tom,

I am sorry to hear that you have troubles with the update.

In such situations we try to recreate the error, create a bug report and notify you once it's fixed.

Could you provide a simple workflow that shows the error?


Not in a place to do this at the moment.  Will get back to you in the next 2 days.

Turns out that I did a Bag of Words prior to these nodes.  So I was going through and changeing 290 thousand documents.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.