Changelog v2.9.0

Changes from v2.8.2 to v2.9.0 (see highlight summary)

Release Date: December 06, 2013

New Nodes:

  • New Node 3857: Send Mail node with html option and file attachments
  • New Node 4741: Save Workflow on execution
  • New Node 4269: Row Filter based on Rule Engine functionality
  • New Node 4270: Row Splitter based on Rule Engine functionality
  • New Node 4531: Cross Joiner
  • New Node 4271: Rule Engine node for variables
  • New Node 4414: Ruleset Editor and Ruleset Predictor with PMML ports
  • New Node 4620: Active Branch Inverter node to toggle activity status of a branch
  • New Node 4399: Data Generation: Counter Generation adds a simple counter to a table
  • New Node 4530: Data Generation: Time Series Generator node
  • New Node 4679: Network Mining: Node that converts a distance matrix to a network
  • New Node 4704: KNIME Labs: JPMML Classifier, Cluster Assigner, and Regression Predictor nodes
  • New Node 3618: SDF Stripper node
  • New Node 4064: Extract Context Properties
  • New Node 4579: New scorer for numerical values
  • New Node 4580: TableColumn to Variable node
  • New Node 4705: Update Weka integration to 3.7
  • New Node 4710: Open Street Map: Geo-Coordinate Row Filter node
  • New Node 4440: Edit Numeric Domain
  • New Node 4441: Edit Nominal Domain (Dictionary)
  • New Node 4442: Edit Nominal Domain
  • New Node 4714: Constant Value Column
  • New Node 4330: List (Remote) Directory
  • New Node 4557: Recursive Loops (with one or two input data tables)
  • New Node 4744: New Pair Extractor node for distance matrices

Enhancements:

  • Enh 3877: Rule Engine node dialog more user friendly: inserts by single click, flow variable access, less confirmation dialogs
  • Enh 3614: Column Filter components should allow for wildcard name filtering (available in different nodes Column Filter, Round Double, Splitter, Collection Create, ...)
  • Enh 4713: Text Mining: Significant speedups due to change of underlying data type and preprocessing on documents (as opposed to bag of words)
  • Enh 4177: Add trusted community update site to default list
  • Enh 4651: Predictor node dialogs looks ugly because of DefaultDialogComponents
  • Enh 4048: TableRow To Variable and TableRow To Variable Loop Start nodes: support missing values and empty table columns
  • Enh 4742: Text Mining: Parallelization of tagger nodes
  • Enh 4757: Text Mining: Word Parser can not read .docx files
  • Enh 4774: Text Mining: Parallelization of Parser nodes
  • Enh 4775: Text Mining: Stopword Filter should provide more build in stop word lists
  • Enh 4275: Scorer node improvements (class ordering, Kappa score)
  • Enh 4522: API for changing default renderers
  • Enh 4616: Make install extensions smarter
  • Enh 4765: Use external browser by default for BIRT report previews
  • Enh 4559: Option in External SSH Tool node to disable host key checks
  • Enh 4649: Loop End and Loop End (2 ports) nodes with new option Ignore empty input tables
  • Enh 4417: CSV Writer Option to enforce unix-style line feeds
  • Enh 4340: Unify predictor outputs
  • Enh 4585: Allow replacing a column in Rule Engine node
  • Enh 3558: Deprecate Variable Based File Reader (can use File Reader with hidden flow variable port instead instead)
  • Enh 3854: Enable Save As... on workflows
  • Enh 4707: Open Street Map: Highlighting of POIs in the Viewer
  • Enh 4711: Open Street Map: preference page to add other tile map services (i.e. maps)
  • Enh 2772: Joiner based on RowIDs (inner join) retain RowIDs in the output
  • Enh 4413: Logistic Regression (Learner) - Allow to change the logit
  • Enh 4553: Java Snippet: Increase length of default Java field names
  • Enh 4554: Java Snippet: Sort list of flow variables in dialog
  • Enh 4555: Java Snippet: Provide more generic methods to work with flow variables in columns
  • Enh 4606: Better joiner default settings (automatic row ID suffix)
  • Enh 4722: Regression (Predictor) does not adhere predictive analysis conventions of the KNIME noding guidlines
  • Enh 4728: Linear regression learner should output model parameters
  • Enh 4596: Text Mining: PubMed Parser uses cached DTD (faster execution & offline parsing of documents)
  • Enh 4567: Network Mining: Optional network inport for Object Inserter node
  • Enh 4760: Network Mining: New feature that adds distance matrix support
  • Enh 4600: Missing Value Handling node dialog to show configuration for columns even if they are no longer present in input
  • Enh 4512: Renderer to Image node doesn't allow to specify output column options (replace or append with new name) - all hardcoded
  • Enh 4551: Unzip node can read from remote location (http://, ftp://, knime://) - used to be only a local path
  • Enh 2224: Linear Regression Node to have 2nd output containing learned parameters
  • Enh 3547: Column Filter components have data type based filter
  • Enh 4594: New Extension Point to allow node vendors to map renamed node factory classes in existing workflows (e.g. com.company -> org.company)
  • Enh 4618: Update web service client library (using Apache CXF 2.7.7 now)
  • Enh 4634: Flow Variable tab in output view to be sortable
  • Enh 4690: Read only support in Local Workspace for data files and meta node templates (eg. meta nodes from the Public Example Server)
  • Enh 4695: Web Service node with more options (output Soap body as XML, more authentication including NTLM, ...)
  • Enh 4698: TableView with new option to change font size
  • Enh 4769: Table Reader node with flow variable control button
  • Enh 4777: Column and flow variable list components support key strokes to jump to next hit
  • Enh 4592: Additional values in Statistics node
  • Enh 3651: Red nodes need to be executable (if fully connected)

Bug Fixes:

  • Bug 4564: ReportBatchExecutor has problems with HTML output
  • Bug 4569: Text Mining: Inputstream encoding problem of DocumentGrabber
  • Bug 4588: GroupBy node adds unknown column to group column list
  • Bug 4676: QuickForm output nodes may not be filled with data when RMI is swapped in web portal (effects web portal users)
  • Bug 4322: Grid Parameter Editor should select field content when stepping through with TAB
  • Bug 4599: PMML writes non standard characters
  • Bug 3686: Variable to TableRow does not recognized new variables
  • Bug 4699: Variable To TableRow node to allow selecting variables in dialog
  • Bug 4624: DenseByteVector accepts only values between 0 and 127
  • Bug 4586: PMML Ensemble Loop End throws IllegalArgumentException when opening dialog (node is not connected)
  • Bug 4672: Delegating Loop End needs to ignore first Spec on outputing last one
  • Bug 4673: JFreeChart Line Chart won't work with color and date/time
  • Bug 4337: QuickForm Text Area Output node uses description as a label
  • Bug 4637: MySQL queries with LIMIT clause don't work in dialogs
  • Bug 4700: Variable To TableColumn node using new Variable Filter Panel
  • Bug 4701: Variable Filter Panel component for dynamic variable selection
  • Bug 4703: Variable to TableRow and to TableColumn nodes with optional Variable in-port
  • Bug 4581: Linear Regression learner should output coefficients and p-values
  • Bug 4685: Text Mining: static singleton tokenizer instance blocking parallel usage
  • Bug 4686: Text Mining: Strings to Document should be parallelized
  • Bug 4687: Text Mining: Stopword Filter can not handle file in drop dir provided as URLs
  • Bug 4755: Text Mining: Pattern in StringsToDocumentCellFactory should be precompiled
  • Bug 4762: Text Mining: PDF Parser fails on PDFs containing images only
  • Bug 4771: Text Mining: Term Grouper fails on document list input table
  • Bug 4772: Text Mining: Parser nodes should not validate selected directory in validateSettings
  • Bug 4590: RowID must not update domain information
  • Bug 3480: SVGCell is not a XMLCell
  • Bug 4383: Disable Server Repository view
  • Bug 4535: ArrayIndexOutOfBounds in Distance Matrix Reader upon empty lines
  • Bug 4536: Update Apache commons math to 3.2
  • Bug 4556: Modular Data Generators should be moved from Labs to Extensions
  • Bug 4576: Coefficient view of PolyReg node occasionally throws ArrayIndexOutOfBoundsException
  • Bug 4589: Upper bounds in table spec wrong when table contains NaN
  • Bug 4622: Disallow placing new categories on top of standard KNIME categories (that is, none should be above I/O)
  • Bug 4715: Remove old image processing nodes (deprecated for 2+ years) from update site
  • Bug 4790: Sparse and DenseBitVector create different hex representations
  • Bug 4798: SparseBitVector creates wrong hex representation for more than 32bits
  • Bug 1070: Regression framework should test save and load
  • Bug 4619: File Reader issues warning for no reason
  • Bug 4756: CSV Reader with bad error message when 'r' as row delimiter is selected
  • Bug 4773: CSV Reader doesn't accept empty delimiter, quote, comment pattern
  • Bug 4804: File Reader doesn't honor limit rows option when guessing settings
  • Bug 4524: Update to JFreeChart 1.0.16
  • Bug 4763: Append (PMML) to all node names with PMML ports (category: Data Manipulation/PMML)
  • Bug 4739: Decision Tree Predictor ignores score attribute in PMML model
  • Bug 4591: StatisticsTables compute wrong maximums when column contains NaN
  • Bug 4764: RegressionPredictor: Output Tablespec mismatch with polynomial models
  • Bug 4767: Linear Regression Learner: Compatibility with JPMML issue
  • Bug 4639: Histogram nodes should handle NaN values
  • Bug 4605: External SSH node should check for absolute paths
  • Bug 4650: Ctrl-OK in dialogs doesn't work with mouse clicks
  • Bug 4604: Missing Value node fails with exection on integer columns (using MIN or MAX replacement)
  • Bug 4400: JFreeChart nodes print too many warning
  • Bug 4708: Open Street Map: Added wrapping of the map in the dialog
  • Bug 4709: Open Street Map: The lat/long bounds in the OSM Map to image nodes are not min/max values
  • Bug 4712: Open Street Map: Size of the image generated from a map is limited
  • Bug 3493: Linear Regression model has numeric precision problems (can't handle singularities & problems with almost constant columns)
  • Bug 4090: Table Creator NullPointerException if column properties get configured before data has been typed into the table
  • Bug 4296: Decision Tree to Image fails without a pruned tree
  • Bug 4386: Bad error message in Java Snippet (e.g. when client code throws NPE)
  • Bug 4439: Logistic Regression Predictor fails when the PMML DataDictionary is inconsistent with learned model
  • Bug 4469: RController must check process return value
  • Bug 4635: Rewrite of Linear Regression node (addresses numerical precison problems, 2nd output with parameters, supports categorical values, PMML using GeneralRegression, etc.)
  • Bug 4721: Regression (Predictor) type mismatch of predicted column
  • Bug 4806: Logistic Regression Learner cannot be canceled in some cases
  • Bug 4689: DialogComponentColumnFilter2 does not call updateModel() after finding a different config in updateComponent()
  • Bug 3477: CSV Reader fails to read files from a flow variable, which specify a path (url works)
  • Bug 4724: File Reader can't handle \r line breaks
  • Bug 3703: New KNIME Molecular Container(!) Type (no need for all Molecule to XYZ converter nodes)
  • Bug 4444: DataColumnProperties.equals may lead to deadlock
  • Bug 4694: Failures writing to table often take 30sec to propagate (for instance duplicate row IDs in file)
  • Bug 4717: class ColumnSelectionPanel does not have a nice baseline (poor layout)
  • Bug 4719: Potential NullPointer in table iterator when memory alert is triggered while a new iterator is created
  • Bug 4746: Settings on nodes may be applied even if the settings don't pass through #validateSettings
  • Bug 4768: Adding/Removing ports from partially connected meta node not possible
  • Bug 4766: List Files node: if only one location is selected, the semi-colon at the end should not be appended
  • Bug 4813: Import workflow wizard sometimes does not enable "Next" button