Changelog for KNIME Analytics Platform 3.5

Changes since 3.5.3

(available as updates only)

Bug Fixes:

  • AP-5555 - Breakpoint dialog not configured correctly
  • AP-9355 - Breakpoint dialog forgets selected flow variable
  • AP-8938 - Twinlist panels of column filters are not of equal width
  • AP-8986 - Send Email node messes up file name of attachments when name contains Japanese characters
  • AP-9097 - "Unzip Files" node dialog permits invalid configuration when writing remote files
  • AP-9101 - Scatter Plot and Line Plot (JavaScript) don't convert old date&time format in representation
  • AP-9105 - Python Script nodes throw EOF Exception when input table contains XML columns with umlaut in XML data
  • AP-9124 - R Integration: Windows build in R distro has broken "tseries" package (requires new dependency to "quantmod")
  • AP-9258 - Feature Selection: End node UI loading issue
  • AP-9325 - Math Formula (Variable) dialog forgets the selected variable on reopen
  • AP-9446 - H2O Integration: Predictor dialogs do not save and load settings correctly
  • AP-9467 - H2O Integration: Cross Validation Loop Start dialog cannot be opened if no nominal column is present
  • AP-9123 - Java Snippet: hidden declarations expand when adding new Input variable
  • AP-9163 - Logistic Regression Learner fails if statistics can't be calculated
  • BD-649 - Spark Normalizer modifies Spark schema of input DataFrame on Spark 2.x
  • BD-650 - Spec mismatch between Spark and KNIME after Spark Collaborative Filtering and Spark Number to Category under Spark 2.x
  • BD-658 - Make JAX-RS 1.1 OSGI bundling more robust

KNIME Analytics Platform 3.5.3

Release date: April 05, 2018

Bug Fixes:

  • AP-9043: Chrome View broken with current Chrome versions
  • AP-9062: Internal Chromium Browser for JavaScript Views does not work anymore with current ChromeDriver
  • AP-8981: Excessive runtime on Windows OS while executing KNIME workflows when "Windows Defender" is turned on
  • AP-8686: Problems with certain Japanese Characters in Java Snippet nodes
  • AP-9035: Race condition in DB connection handling
  • AP-8265: REST nodes not supporting duplicate column handling
  • AP-9027: JavaScript boxplot should report statistics consistent with Java boxplot
  • AP-9002: Selection in the tag cloud view does not work properly after restoring a workflow
  • AP-8975: Window Loop Start creates empty table when first window empty and second contains rows
  • AP-8962: Document Grabber cuts text when superscript occurs
  • AP-8960: Logistic Regression Learner node with error "DataSpec generated by configure does not match spec after execution"
  • AP-8956: "Zip Files" node dialog permits invalid configuration when writing remote files
  • AP-8915: Date&Time nodes do not work with some locales
  • AP-8896: Zip File node shows success after failing to write to remote location
  • AP-8894: H2O Integration: Kmeans fails, if columns are excluded for training
  • AP-8887: Writing remote files with non-ASCII characters in their name doesn't work
  • AP-8862: Table Writer fails when writing large tables to remote locations
  • AP-8783: Text Mining: Dictionary Tagger problem with multiple words
  • AP-8749: Excel nodes disappear when opening workflow
  • AP-8701: Pivoting node returns empty table if pivoting column contains missing values and the ignore missing values option is enabled
  • AP-8578: H2O Integration: PCA model does not work properly if written out and read in again
  • AP-8577: H2O Integration: Column Selection does not save Wildcard/RegEx and Type Selection
  • AP-8388: Topic Extractor (Parallel LDA): Problems with warning messages
  • AP-8231: (File) Reader dialog doesn't open if there are special characters in paths
  • AP-7882: H2O Integration: Enforce input column type in H2O Table to Frame node
  • AP-7387: Javascript Lift Chart: should be able to specify chart title and subtitle
  • AP-5607: Double clicking tab (with workflow name in) doesn't maximise the tab to fill the window
  • BD-634: (Big Data Extensions): Mass column renaming on Spark 2.x takes too long
  • BD-537: (Big Data Extensions): Hive Loader fails with non-default database
  • BD-448: (Big Data Extensions): Spark SVD fails if compute U matrix option is not selected

KNIME Analytics Platform 3.5.2

Release date: February 06, 2018

Enhancements:

  • AP-8642: "Nightly Build" warning with toggle to hide/show the dialog
  • AP-8098: Validation check during the execution of the Slider Input

Bug Fixes:

  • AP-8658: KNIME Logger swallows all messages except those coming from KNIME (org.knime / com.knime)
  • AP-8640: Python: node descriptions inconsistent
  • AP-8763: JavaScript View nodes don't persist their state correctly
  • AP-8759: Output nodes don't persist their state correctly
  • AP-8755: Quickform nodes with input table don't persist their state correctly
  • AP-8699: Dictionary Tagger Incorrectly Assigns Tags
  • AP-8606: H2OGLRMNodeModel uses local implementation of H2OSession (H2OFileImportSession)
  • AP-8530: Conditional Box Plot (JS) node raises NullPointerException, if category columns have no domain values
  • AP-8510: x-column in Conditional Box Plot (JS) is always named "class"
  • AP-8499: Data Explorer: small histograms when large numbers of columns present
  • AP-8051: Javascript Network Viewer node font setting not working
  • AP-7214: JavaScript Line Plot: dialog allows Selection mode to be enabled, but the view itself doesn't seem to support selection
  • AP-7104: Text Mining: relative TF of terms of tagged documents may not sum up to 1
  • AP-7068: Javascript Stacked Area Plot: tooltips not updating when the chart type is changed

KNIME Analytics Platform 3.5.1

Release date: December 21, 2017

Enhancements:

  • AP-8583: R Nodes: Longer timeout (30s) to connect to R process
  • AP-8567: Update Marvin extension to 3.5.0.v0800
  • AP-8502: DL Keras: Enhance View
  • BD-425: (Big Data Extensions): Allow setting the request timeout for the Jobserver REST connection

Bug Fixes:

  • AP-8535: New in 3.5: Java runtime not accepted by some MacOS
  • AP-8580: DL Keras: Avoid negative times in view
  • AP-8536: Very slow startup after update from 3.4 on some installations
  • AP-8520: Python Arrow: Temporary files do not get deleted on windows
  • AP-8464: DL Framework: Allow execution of networks trained with custom loss function
  • AP-7755: Metanode reconfigure port types dialog has incorrect size (Linux)
  • BD-605: (Big Data Extensions): Fix wrong output spec in Spark 2.0 sum aggregation
  • BD-426: (Big Data Extensions): Spark jobs should never time out

KNIME Analytics Platform 3.5.0

(see highlight summary)

Release date: December 06, 2017

New Nodes:

  • AP-7925: Google Sheets Interactive Service Provider
  • AP-7550: Google Sheets Reader
  • AP-7551: Google Sheets Writer (all in all 3 nodes)
  • AP-7043: OPTICS distance based clustering
  • AP-4931: Math Formula (Multi Column)
  • AP-5208: Windowing Loop Start node
  • AP-7546: JavaScript Data Explorer
  • AP-7562: JavaScript Table Editor
  • AP-5173: JavaScript based Tag Cloud
  • AP-7320: Run R Model in Microsoft SQL Server
  • AP-7508: H2O Integration: Naive Bayes
  • AP-7506: H2O Integration: GLRM
  • AP-7460: H2O Node: Scorer
  • AP-7449: H2O Mojo Plugin: Implement H2OMojoPortObject
  • AP-7448: H2O Mojo Plugin: Mojo Reader
  • AP-7447: H2O Integration: Mojo Exporter
  • AP-7446: H2O Mojo Plugin: Mojo Predictor (Classification and Regression)
  • AP-7445: H2O Integration: PCA
  • AP-7444: H2O Integration: K-Means
  • AP-7443: H2O Nodes: GLM Learners
  • AP-7438: H2O Integration: Column Filter
  • AP-8053: H2O Mojo Plugin: Mojo Predictor (Clustering, Auto Encoding, Dimension Reduction, Word Embedding)
  • AP-7758: H2O Integration: GLRM node which directly imputes missing values
  • AP-7731: H2O Integration: PCA Node (refactor)
  • BD-539: (Big Data Extensions): Spark GroupBy node
  • BD-224: (Big Data Extensions): Spark Missing Value node

Enhancements:

  • AP-8522: KNIME Deep Learning - Keras Integration
  • AP-7623: R Binaries on Windows updated to version 3.4.2 (was: 3.0.3)
  • AP-7878: Updated bundled JRE to 1.8.0_152
  • AP-7876: ChemAxon / Infocom Extension updated to 3.4.2.v0701
  • AP-8251: Add Gradient Boosted Trees and Simple Regression Tree support to generic PMML Predictor
  • AP-8249: TableView with 'Ctrl- +/-' to increase or decrease font size
  • AP-8224: Python (Labs): Support lazy data loading during Python code execution
  • AP-8184: Add example field to JSON Output node
  • AP-8183: Add description field to JSON Input and JSON Output nodes
  • AP-8162: H2O Mojo Plugin: Minor improvements
  • AP-8090: H2O: New colors for the H2O ports
  • AP-8128: Duration to Number: Make clear that only time-based durations can be processed
  • AP-8119: PMML Support for Gradient Boosted Trees and Regression Trees
  • AP-8093: Scan workflows for quickform specifications and generate OpenAPI definition
  • AP-8082: Microsoft SQL Server Connector to use official SQL Server Driver (com.microsoft.sqlserver)
  • AP-8075: Case insensitive matching of file name extensions for drag & drop operations from KNIME explorer (= drag file to workbench)
  • AP-8033: DeepLearning4J: Update to 0.9.1
  • AP-7982: Workflow Loader to show confirmation dialog if a workflow created with a 'future' version is loaded
  • AP-7977: Make colors of H2O port objects easier to distinguish
  • AP-7959: Show error messages in metanode view and WebPortal for unavailable view nodes
  • AP-8363: Python: Better error message for Collections in Collections
  • AP-7937: Python: Receive log messages asynchronous to job execution.
  • AP-7936: Python: Kernel should give client control over logger output
  • AP-7932: Python: Improve kernel startup time
  • AP-7865: Python (Labs): PythonKernel: Make chunk size configurable
  • AP-7863: Python: Better exception messages on failure
  • AP-7721: Python Flatbuffers: Faster deserialization of arrays
  • AP-7859: Promote JavaScript View nodes out of KNIME Labs
  • AP-7857: Single Selection in JavaScript Table View
  • AP-7846: Promote KNIME REST nodes from KNIME Labs
  • AP-7827: New usage configuration dialog for wrapped metanodes
  • AP-7806: Save OpenAPI definition to workflow directory on workflow save (used in KNIME Server's SwaggerUI)
  • AP-7802: Distributing Chromium as part of a standard KNIME product build on Windows (used for JS Views)
  • AP-7800: Remove probability options from Regression Predictor
  • AP-7774: "Quick Form (legacy)" deprecated -- fully replaced by new JavaScript based Quick Forms
  • AP-7676: Feature Elimination Nodes (Labs) promoted to standard KNIME category
  • AP-7654: (API) Expose Weka Port Object package in plugin.xml
  • AP-7303: Add percentages to occurrences output of statistics node
  • AP-7047: Make checkbox always aligned with its label in Boolean Input Quickform
  • AP-6943: String Input Quickform should provide choice to use textarea for longer user input
  • AP-6906: Usability of Call Workflow Nodes with Respect to Input and Output Parameters
  • AP-7965: Text Mining: Streamline TIKA Parser
  • AP-7767: Text Mining: Strings to Document to support empty titles
  • AP-8194: Text Mining: StanfordNLP NE Learner to support multi words in dictionary
  • AP-8192: Text Mining: Adjust options for Term to Structure node
  • AP-7488: Text Mining: (API) Remove bundling of libraries in /lib folder (e.g. guava)
  • AP-7371: Text Mining: Bag of Words Creator: Add option to carry over additional columns from source table
  • AP-7324: Text Mining: Tagger nodes to retain all input columns (were removed in previous versions)
  • AP-7323: Text Mining: Tagger nodes throw errors on input tables with >1 document columns
  • AP-6813: Text Mining: Document Vector Adapter node should have model input
  • AP-6753: Text Mining: Document Vector node should have model out port
  • AP-4210: R Views supporting SVG image output
  • BD-500: (Big Data Extensions): Support regression learning in Spark Decision Tree Learner node
  • BD-451: (Big Data Extensions): New option for Create Spark Context node to destroy context on dispose
  • BD-352: (Big Data Extensions): Azure Blob Store support for Hive Loader
  • BD-347: (Big Data Extensions): Amazon S3 support for Hive Loader
  • BD-152: (Big Data Extensions): Update open-source Hive JDBC drivers to v1.1
  • BD-103: (Big Data Extensions): Update Hadoop libraries to v2.7.4

Bug Fixes:

  • AP-8277: Cannot save changes to wrapped metanode containing a Range Slider filter definition node
  • AP-7649: Download/Upload nodes don't support KNIME protocol on KNIME Server any more
  • AP-8457: Details text for python integration confusing
  • AP-8450: Javascript views don't save the view changes, if inside of a wrapped metanode
  • AP-8430: Mac: Tooltip on long node warning messages may kill KNIME app (and whole Mac session also)
  • AP-8414: Text Mining: StanfordNLP Scorer fails on Linux/Mac
  • AP-8404: Workflow Coach: Double-check that is appears without community stats
  • AP-8398: Text Mining: Document Vector Apply has ports in wrong order (model should come first)
  • AP-8397: Problem setting path to R
  • AP-8352: Python: Provide error message in node, not only console
  • AP-8342: DeepLearning4J: Memory error should lead to better error message
  • AP-8329: Python: Kernel start up in nodes hangs if the default python executable can't be found
  • AP-8328: Python: Preference page resets default to Python 2 when pressing apply
  • AP-8279: The Amazon S3 Connector node fails with access denied error without S3 list bucket permission.
  • AP-8272: Python (Labs): streetnames not properly migrated
  • AP-8269: New DateDialogComponent does not handle invalid input
  • AP-8245: JavaScript Table View: enable selection by default
  • AP-8225: Rule-based Row Filter (Dictionary) doesn't recognize changes in dictionary table
  • AP-8220: Python (Labs): Kill Python Process on Cancel
  • AP-8218: Python (Labs): NPE when dialog is closed before python setup is completed
  • AP-8188: Text Mining: Fix missing value/empty table handling
  • AP-8182: Python (Labs): Include modules extension point in typextensions plugin
  • AP-8171: Marking of wrapped metanodes is slow
  • AP-8120: PMML Transformation cannot handle Missing Values for Boolean Cells
  • AP-8111: Text Mining: AIOOBE in Meta Info Extractor node
  • AP-8091: Text Mining: Topic Extractor (LDA) fails on missing values and empty table
  • AP-8081: Python (Labs): Non-consecutive RowIDs in table produced by Python Source (Labs) for large numbers of rows
  • AP-8077: Auto-install of textprocessing integration with DeepLearning4J doesn't work properly
  • AP-8076: DeepLearning4J: Word Vector Apply node inconsistencies
  • AP-8055: JoinTableIterator leaks file descriptors
  • AP-8054: Random Number Assigner takes the value from the row above as min and max value
  • AP-8008: Data Explorer: histograms only display "correctly" on later pages
  • AP-7943: Python CSV Serialization fails on Windows for Longs
  • AP-7866: Python Arrow: Buffer allocation fails with large data types
  • AP-7864: Python Arrow: Make sure temporary files are deleted on failure
  • AP-7854: Text Mining: AIOOBE in Document Data Extractor
  • AP-7773: Google API connector saves filepath instead of information about file
  • AP-7772: Google API Connector does not support KNIME Protokoll
  • AP-7730: Text Mining: Strings To Document creates empty output if selected column does not exist instead of setting a warning
  • AP-7724: Text Mining: StanfordNLP NE Scorer can not be configured
  • AP-7691: Text Mining: Document Vector node does not reset Row ID count after reset
  • AP-7588: Text Mining: String to Document throws cryptic error message on invalid settings
  • AP-7581: Javascript Table View: "select all" should not select rows that aren't in the filter
  • AP-7468: NodeExecutionResult doesn't save flow variables pushed by the node
  • AP-7455: Workflow Coach retrieves wrong data during update if connection is redirected
  • AP-7399: Text Mining: Strings to Document node fails on data table that already contains a column called "Document"
  • AP-7388: R to PMML fails since R uses PMML 4.3
  • AP-7178: REST nodes cannot connect to server requiring SNI
  • AP-4457: Text Mining: Frequency Filter should not require min and max threshold
  • BD-530: (Big Data Extensions): Race condition in jobserver rest client with authentication
  • BD-513: (Big Data Extensions): Destroy Spark Context throws HeadlessException on server if confirmation dialog is enabled
  • BD-444: (Big Data Extensions): Spark Joiner: Filter duplicates option leads to IllegalArgumentException
  • BD-433: (Big Data Extensions): Spark Java Snippet freezes GUI while preparing autocompletion
  • BD-351: (Big Data Extensions): Delete only RDDs that belong to an open Spark context
  • BD-51: (Big Data Extensions): Parallel database operations on Hive/Impala fail with "out of sequence response"