Lhasa community contribution changelog

21/05/2017

  1. Release 3 new KNIME nodes
    1. Kill Configure: when eneabled returns a null table spec from the configure method
    2. Remove Empty Rows: removes rows where all cells are empty
    3. String Cell to File: writes a file for each cell in specified column.

19/11/2016

  1. Bug with multi performance and binary scorer fixed for KNIME 3.1, 3.2, 3.3 and trunk releases. The bug identified occurs when the target column does equal either the active or inactive value specified. Fix provides removes these row's from the performance measures. 

18/07/2016

1. With the release of the new CDK jar the knew SMARTCyp updated have been merged into the 3.2 build (see changelong: 06/05/2016).

09/07/2016

1. KNIME 3.2 branch released

06/05/2016 - Replace SMARTCyp node with new version - nightly

  1. Placing an IAtomContainer into a CDK cell causes atom renumbering. Highlight properties are jumbled and a connection with the input structure atom numbering (if present) is lost
  2. The SMARTCyp node now directly passes the MolValue and SdfValue cells to SMARTCyp 2.4.2 application. Atom numbering for these formats is retained. 
  3. CDK Adapter cells are now only used for cell types other than MolValue and SdfValue
  4. Node no longer outputs the input structure in each row of the detailed output table. 

05/05/2016 - Upgrade to existing KNIME nodes for version 3.1 - nightly

  1. Where possible the existing general nodes implement the streaming API
  2. SMARTCyp and WhichCyp can now generate SVG or PNG using the new CDK Depiction API
  3. SMARTCyp provides the option to output the atom ID under 0 based index. This allows the use of the value in downstream highlighting without decrementing the atom ID. The default is false so expected behaviour should not change.

16/02/2016 - Metabolism nodes released to KNIME 3.1

  1. Released SMARTCyp 2.4.2 and WhichCyp 1.2 as KNIME nodes
  2. Minor code cleanup of some generic nodes

16/10/2015 - upgrade to KNIME v3.0

  1. Updated various NodeModel classes to use the new BufferedDataContainer#size() method replacing the now deprecated BufferedDataTable#getRowCount() method.
  2. Updated the TableToHtml node to use a StringBuffer for building the HTML String. You are unlikely to notice any performance improvement. 

Initial release:

Released the following nodes:

  • Discretise
  • Binary scorer
  • Binned performance
  • Result
  • Dumb Joiner
  • Row Splitter (col+)