Changelog v2.0.2

Changes from v2.0.1 to v2.0.2

Release Date: April 7, 2009

This is a bug fix release, which addresses the following issues:

  • Bug 1677: Strange error message on feature elimination filter
  • Bug 1680: "Manhattan" norm in Hierarchical Cluster node is actually Euclidean
  • Bug 1683: ARFF reader can't handle empty nominal values
  • Bug 1684: getFloat(String key) is missing in ConfigRO
  • Bug 1688: SD Reader and Extractor nodes die if the SD format is incorrect
  • Bug 1689: SdfValue API must require the "$$$$" as last line
  • Bug 1692: Copying nodes with non-deletable connections, makes copied nodes also non-deletable
  • Bug 1693: Sdf reader changed compound ID
  • Bug 1694: FileReader: once the delimiters are set they cannot be changed back
  • Bug 1695: SDFWriter writes missing properties as '?'
  • Bug 1699: PDF/HTML Generator nodes don't produce images
  • Bug 1700: PDF/HTML Generator nodes throw NPE on invalid file location
  • Bug 1705: RowFilter node ignores "case sensitive match" flag
  • Bug 1706: DriverManager#setTimeout() not implemented by all database driver
  • Bug 1707: Dendogram inconsistent order of data points after re-load
  • Bug 1710: DefaultNodeSettingsPane misses loadAdditionalSettings(NS, DTS[]) method
  • Bug 1712: SDF Reader Dialog possibly throws ConcurrentModificationException
  • Bug 1715: Possibly version conflict when loading 2.0.1 workflows with 2.0
  • Bug 1716: Weka node "SerializedClassifier" unusuable
  • Bug 1719: FileReader should created ColName for empty col header
  • Bug 1724: Numeric Binner precision restricted to 0.###1
  • Bug 1737: Sdf and Mol2 reader have memory problems (row keys reference entire record)
  • Bug (no case): Scorer uses input hilite handler also at output; it shouldn't
  • Bug (no case): BasicPlotter.interpolateLine does wrong cast
  • Feature Request 1730: Chemistry MolCell should also implement SdfValue
  • Feature Request 1731: Need more debug messages when dirs in workflow are created/removed
  • Feature Request 1728: Cells to be held in memory should be customizable (using property in knime.ini)
  • Feature Request (no case): R Local nodes support row IDs