Changelog for KNIME Analytics Platform 4.0

KNIME Analytics Platform 4.0.1

Release date: August 15, 2019

Enhancement

  • AP-12371: Display technical JDBC driver ID together with human-readable driver name
  • AP-12119: Copy&Paste of workflow from Hub
  • AP-12114: Allow to disable DB transactions
  • AP-12001: UI improvements to the metadata editor

Bug Fixes

  • AP-12458: Dragging on a shared component containing a credentials configuration from remote freezes KNIME
  • AP-12398: Master Page panel in BIRT is inaccessible (greyed out)
  • AP-12357: Order of possible values produced by File Reader is non-deterministic
  • AP-12341: Excel Reader - More Options dialog grayed out
  • AP-12239: Joiner node with empty table in top port generates ArrayIndexOutOfBoundsException in 4.0.0
  • AP-12238: KNIME 4.0 ignores logging information from any class except {org|com}.knime.* (breaks log messaging from partner extensions)
  • AP-12198: Chromium 75 update causes error modals to automatically close
  • AP-12159: AP does not prompt for initial workspace location on first startup
  • BD-950: (Big Data Extensions): Create Spark Context (Livy) fails with UnsupportedOperationException when using S3 Connection as file system
  • AP-12545: Endless update of annotation data if workflow is set dirty
  • AP-12542: Opening workspace created with AP 3.5 yields nullpointer
  • AP-12525: Violin Plot View fails with JS error if "Skip Rows with missing values" is selected
  • AP-12510: SHAP Loop Start can get stuck in sampling loop
  • AP-12501: DB Loader fails if Hive or Impala table does not exist during configure
  • AP-12496: Send to Tableau Server (Hyper): Cannot set project id
  • AP-12494: Improve the validation mechanism of database connector nodes
  • AP-12484: XGBoost Predictor can't deal with different column orders
  • AP-12478: R: Error logged: Error reading output of R process
  • AP-12470: SSH Connection does not correctly handle remote knime:// URLs for private key
  • AP-12459: DeepLearning4J: Settings loading dependent on locale
  • AP-12451: Hive/Impala Loader fails if default schema is read-only
  • AP-12373: Prevent registering JDBC drivers with built-in driver IDs/names
  • AP-12369: DB driver URL template should be used by all dedicated DB connector nodes when driver is selected
  • AP-12364: DB connection context property resolution does not work with Kerberos
  • AP-12333: Cannot move/copy and overwrite multiple workflows in KNIME Explorer
  • AP-12329: Parameterized DB Query Reader fails if the same column name is used multiple times in the query
  • AP-12282: Data files getting overwritten without being asked when drag'n'dropped or copied&pasted into explorer
  • AP-12275: Range Slider Filter node outputs rounding erros
  • AP-12270: H2O Integration: Add validation setting to GLM to fix lambda search
  • AP-12242: H2O Integration: Partitioning and missing value imputation may change the column types
  • AP-12232: KNIME Hub cannot be removed via preferences
  • AP-12225: Metanode/component port migration errors
  • AP-12223: Violin Plot fixes
  • AP-12221: Bubble Chart node displays different bubble size if group-by column is provided
  • AP-12220: Drag'n'drop problem with Image Processing nodes from Hub to AP
  • AP-12216: When clicking "deploy to server" while the server is still fetching the view will never update
  • AP-12209: Plotly (Scatter Plot) no data in image when using OpenGL library
  • AP-12207: List Remote Files can't handle knime:// protocol with whitespaces
  • AP-12201: Variable Expressions shows wrong label for the evaluation
  • AP-12200: H2O Generalized Linear Model Learner nodes fail if lambda = 0
  • AP-12197: Component not found on Hub even though it exists
  • AP-12181: Metanode dialogs refer in disclaimer text to 'quickforms' and 'components' -- they shouldn't.
  • AP-12180: Plotly Bubble Plot: leaving title blank still results in a default title being displayed
  • AP-12179: Divide by zero error in Statistics node
  • AP-12164: Starting KNIME AP on Windows with new workspace yields all mount points fully expanded
  • AP-12162: Fix Plotly suggested extension or update how nodes are registered with their extensions
  • AP-12161: File Upload Widget does not behave like File Upload
  • AP-12160: Upload node truncate prefix function does not work for prefixes containing whitespaces
  • AP-12157: Component referred to as Meta Node when sharing of template fails
  • AP-12147: Document Grabber Node gives n-1 documents
  • AP-12101: Table View: Multiple Scroll bars
  • AP-12100: Pie/Donut Chart: Multiple Scroll bars
  • AP-12099: Tagged Document Viewer: Multiple Scroll bars
  • AP-12097: PDP Obsolete Axis Title Settings
  • AP-12095: Node Description PDP/ICE node does not describes correctly input ports
  • AP-12093: PDP/ICE node description should reference a soon to be published Hub example workflow
  • AP-12061: Naming and description of Out-of-Bag Port of Tree Ensemble and Random Forest learners is confusing
  • AP-12060: Plotly nodes: Create legend group for missing values
  • AP-12047: Table Validator typo in error message "CONVERTION"
  • AP-11990: Jumping workflow when editing node annotation
  • AP-11567: RServe instances dies during execution
  • AP-11310: Credential Input nodes always use first workflow credential if multiple credentials are available
  • AP-11224: Explorer: Error while moving directory to parent directory
  • AP-11150: Components: inconsistent tab-title and node annotation
  • AP-7245: Gradient Boosted Tree ArrayIndexOutOfBounds problem
  • AP-6372: Parallel Chunk loop with large parallelization and nested scopes might error with: "Too many nested loops: 128"
  • BD-957: (Big Data Extensions): Create Local Big Data Environment should always use the built in driver
  • BD-956: (Big Data Extensions): Parquet Reader's type by name mapping dialog does not work
  • BD-955: (Big Data Extensions): Parquet Reader/Writer should support Parquet types ENUM, JSON, INT8/16/32/64, TIME_MICROS and TIMESTAMP_MICROS
  • BD-951: (Big Data Extensions): Spark Job Server Job Controller fails on job result deserialization with tmp file exists
  • BD-942: (Big Data Extensions): PySpark nodes only update FlowVariables if Dialog is opened
  • BD-938: (Big Data Extensions): DB Loader with Impala fails to upload tables with LocalDate(Time) columns
  • BD-937: (Big Data Extensions): DB Loader with Hive fails on CDH 6 with ArrayIndexOutOfBoundsException
  • BD-936: (Big Data Extensions): DB Port of Create Local Big Data Environment node fails to load when opening an already executed workflow
  • BD-935: (Big Data Extensions): PySpark configuration dialog does not open due to NullPointerException when Spark 1.x is chosen in Spark preferences

KNIME Analytics Platform 4.0.0

(see highlight summary)

Release date: June 27, 2019

New Nodes:

  • AP-11517: Split Quickforms into Configuration and Wizard Nodes
  • AP-12154: Plotly Visualization Extension (12 new visualization nodes using plotly.js -- Radar Plot, 3D scatter plot, error bars plot, bubble chart, etc.)
  • AP-11848: Amazon Authenticator
  • AP-11755: Integration of AWS Machine Learning Services
  • AP-11721: SHAP Loop (loop construct for determining feature importance for individual predictions)
  • AP-11594: LIME Loop Start (performs sampling for Local Interpretable Model-agnostic Explanations)
  • AP-11471: DL: Create Executors for different backends
  • AP-10983: Container Input (Row)
  • AP-11319: Container Output (Row)
  • AP-11292: ONNX to TensorFlow converter node
  • AP-11071: T-SNE Node (popular dimensionality reduction technique)
  • AP-10903: H2O Isolation Forest (anomaly detection based on H2O Isolation Forest)
  • AP-10674: LDA Apply
  • AP-10214: Text Mining: Unique Term Extractor node
  • AP-9901: Duplicate Row Filter
  • AP-9852: JS Partial Dependence and ICE View (visualizes feature impacts on model predictions)
  • AP-9486: Document Viewer (JavaScript)
  • AP-9485: Text Processing: Brat Document Writer
  • AP-9272: DB Loader
  • AP-8687: Value Filter Definition Quickform node
  • AP-6993: Text Mining: Relation triplet extractor
  • BD-866 / BD-867 / BD-870: (Big Data Extensions): Spark Decision Tree/Random Forest/Gradient Boosted Trees Learner (based on Spark ML)
  • BD-914: (Big Data Extensions): Spark Predictor (Classification)
  • BD-926 / BD-927 / BD-928: (Big Data Extensions): Spark Decision Tree/Gradient Boosted Trees/Random Forest Learner (Regression)
  • BD-929: (Big Data Extensions): Spark Predictor (Regression)
  • BD-882: (Big Data Extensions): Create Local Big Data Environment (new)
  • BD-747: (Big Data Extensions): DB to Spark (new)
  • BD-746: (Big Data Extensions): Spark to DB (new)
  • BD-745: (Big Data Extensions): Spark to Impala (new)
  • BD-744: (Big Data Extensions): Spark to Hive (new)
  • BD-743: (Big Data Extensions): Hive to Spark (new)
  • BD-742: (Big Data Extensions): Impala to Spark (new)
  • BD-737: (Big Data Extensions): Impala Connector (new)
  • BD-736: (Big Data Extensions): Hive Connector (new)
  • BD-335: (Big Data Extensions): Spark Repartition

Enhancements:

  • AP-12059: Update Infocom/ChemAxon extension to 4.0.0.v191300
  • AP-12032: Layout Editor: Quickform should be replaced with Wizard node
  • AP-11999: Prevent stop-the-world garbage collection when adding rows to buffer
  • AP-11931: Redesign Welcome Page
  • AP-11913: Improve error message for ONNX to TensorFlow Network Converter node
  • AP-11879: Tableau Server: Add support for authentication via API token
  • AP-11858: Revisit Garbage Collection in KNIME
  • AP-11852: Enable asynchronous writing of tables to disk at loop end nodes
  • AP-11851: Add suggestion in Windows installer to add KNIME AP as an exception for Windows Defender
  • AP-11814: Action to open linked components in the browser from AP
  • AP-11788: Link (and update) templates/components when dragged and dropped from browser
  • AP-11777: Provide Plotly example templates for the Generic JS view
  • AP-11773: Parameter Optimization Loop: Bayesian Optimization with Tree-structured Parzen Estimator (TPE)
  • AP-11769: Add option for a custom error message to Breakpoint
  • AP-11767: Add MAPE error metric to the Numeric Scorer
  • AP-11764: Improve error message display of Wrapped Metanodes
  • AP-11763: Enable templates in Generic JS View
  • AP-11738: Search View for Hub
  • AP-11733: Retire Wrapped Metanodes; Introduce "Components"
  • AP-11722: DevDoc: Add .qualifier to version number in NumberFormatter Example
  • AP-11713: Various workflowalizer improvements
  • AP-11701: Keras: Add causal as padding option for Conv1D
  • AP-11650: Kafka Producer: Add option to send records asynchronously
  • AP-11635: Support sending of JDBC parameter via JDBC url and not only properties file
  • AP-11628: Improve workflow metadata editing
  • AP-11575: Tweak dictionary page size for Parquet Table Store
  • AP-11424: Shapley Values Loop (loop construct to determine feature importance for individual predictions)
  • AP-11365: Prevent unnecessary db query execution between configure and execute
  • AP-11336: Support the extended character limit also for Retweets
  • AP-11334: Implement a sliding window cache table
  • AP-11277: CSV Writer with option to write numbers in scientific notation (default was: scientific but will be non-scientific now)
  • AP-11244: Upgrade Java Version (from Oracle JRE 1.8.0_152 to OpenJDK 8u202-b08)
  • AP-11221: Add support in S3 node for switching roles
  • AP-11217: Unit-Test-based Performance Benchmarking Suite for KNIME Core
  • AP-11202: Move domain update into separate thread
  • AP-11191: Remove Warning "Table has k rows, ignored all rows except the first one" from Table row to variable node
  • AP-11165: Less verbose logging in knime.log file and KNIME console ("WARN" as new default minimum level) -- speed-up for logging intensive workflows, e.g. using loops
  • AP-11091: Replace GZIP compression of KNIME tables with speed-optimized algorithm
  • AP-11076: Revise Kerberos preference page
  • AP-11044: Tensor Processing: Explicitly handle forks in graph during execution
  • AP-11008: Veto annotation selection when clicking out of content edit mode
  • AP-10887: Start annotation content edit in annotation editing mode by single-click on selected annotation
  • AP-10871: Text Mining: Meta Info Extractor to fit to new Meta Info Inserter
  • AP-10870: Text Mining: Meta Info Inserter to add multiple meta information
  • AP-10848: Add VirtualTensorOpSpecs for all "Views." operations from ImgLib
  • AP-10844: Allow FileStoreCells referencing zero FileStores
  • AP-10685: Internal data handling: Optimize asynchronous table write
  • AP-10684: Optimize when and how to write tables to disk
  • AP-10682: Text Mining: Add option for case sensitivity for StanfordNLP NE Learner
  • AP-10546: Image Processing: LabelingValue and Cells
  • AP-10534: DL Framework: Support String inputs
  • AP-10499: Text Mining: Improve preprocessing nodes that use regex
  • AP-10367: Random Strategy for Feature Selection
  • AP-10340: Genetic Algorithm for Feature Selection
  • AP-10215: Text Mining: Option for Dict Replacer (2 inports) to replace terms with default value that are not included in dictionary
  • AP-10205: Column-based storage: Row iterator for range of rows
  • AP-10139: Extract Date&Time Fields: Add option to extract the week-based year
  • AP-10082: Refactor LDA Node
  • AP-9903: Range slider should be able to fix one of the 2 ends so you filter by just moving lower or upper bound.
  • AP-9606: Dropping file from Explorer should mimic node-from-Node-Repository drop behavior
  • AP-9574: Enable focus traversal in node dialogs (via TAB, SHIFT-TAB, CTRL-TAB and CTRL-SHIFT-TAB)
  • AP-9479: REST Client Nodes (GET, POST, ...) support NTLM authentication
  • AP-9129: Complement context menu settings with toolbar for annotations
  • AP-8904: DL Framework: Deprecation mechanism for converters
  • AP-8505: DL Python: Preference page
  • AP-8465: Python: Better error message if arrow is not installed or installed in a wrong version
  • AP-8433: Python Installation: Make it easier to setup KNIME Python
  • AP-8308: DL: Allow to select used CPU/GPU in DL nodes
  • AP-7816: Text Mining: constructor of Author class must not set different name values other than given parameters
  • AP-7019: Support knime URLs for 'Binary Objects to Files'
  • AP-5943: Add encoding options to SDF Reader and Writer
  • AP-5320: Google Analytics Query Node - Improved segments support (migrate to new dynamic segments API, add advanced segments support)
  • AP-5087: Number to String and String to Number (with and without PMML) with new column filter panel ("enforce in/exclusion")
  • BD-336: (Big Data Extensions): Show number of partitions in tooltip and view of Spark data output ports
  • BD-909: (Big Data Extensions): Remove license check on KNIME Server for Parquet/ORC Reader and Writer nodes
  • BD-740 (Big Data Extensions): Hive support for DB Loader
  • BD-741 (Big Data Extensions): Impala support for DB Loader
  • BD-795: (Big Data Extensions): Code completion in PySpark nodes

Bug Fixes:

  • AP-11998: OSM nodes: Default Open Street Map Tile-Server not available anymore
  • AP-11994: "Quick jump to function" does not work properly for String Manipulation/Math Formula
  • AP-11920: Expression syntax check broken in some nodes
  • AP-11285: Failed attempt to create multiple SWT Displays
  • AP-12073: Dynamic Node Generation Framework logs Error for SVG outport, even if SVG image is not output.
  • AP-12068: Can't use "Configuration" and "Widget" nodes in a sequence
  • AP-12036: Window for Shared Components Link Type still references "Metanode Template"
  • AP-12002: Deadlock when opening a workflow with an executed node that wants to access the flow variables via NodeModel.getAvailableInputFlowVariables() in NodeModel.loadInternals()
  • AP-11992: Wrong Message when Dragging'n'dropping components from browser (hub) to AP workbench
  • AP-11961: Reading and Writing BLOB/Binary object columns doesn't work for various databases
  • AP-11950: Allow to select no constant column in Feature Selection Loop Start nodes
  • AP-11917: PCA Apply node doesn't execute if number of dimensions > number of input columns even if the number of dimensions selection is disabled
  • AP-11882: Sharing Components whose name contains trailing spaces fails
  • AP-11877: Python: Nodes not printing to dialog console
  • AP-11867: H2O Integration: MOJO Reader kills JVM if the MOJO is overwritten during reading
  • AP-11854: DoubleCell hashCode not compatible with equals for 0.0 and -0.0
  • AP-11836: Location of File in Excel Reader cannot be controlled via FlowVariable
  • AP-11798: Example workflows in LOCAL contain deprecated nodes
  • AP-11784: H2O Integration: Rename H2O Regression Scorer to H2O Numeric Scorer
  • AP-11783: H2O Integration: MOJO predictors don't set warnings about missing columns during execution
  • AP-11780: Workflow Migration Tool throws NullPointerException when migrating nodes with lazily initailized NodeFactories
  • AP-11778: Workflow Migration Tool "swallows" unconnected ports of metanodes
  • AP-11758: Container Output (JSON) always takes first column to fill example
  • AP-11752: Expression Engine: Cannot return collections
  • AP-11747: Buffer flusher persists beyond workflow execution
  • AP-11744: Forward compatibility: Enable the loading of workflows saved with snappy-compressed data in pre-3.8.0 KNIME AP
  • AP-11725: H2O Integration: Rework Generalized Linear Model Learner dialog
  • AP-11710: Resetting Wrapped Metanode keeps successors in queued state
  • AP-11705: Favorite Nodes View: Most frequently and last used nodes won't get updated when nodes are dragged and dropped
  • AP-11667: Breakpoint node remains active even if in inactive branch
  • AP-11655: Database Driver window missing Dialog buttons on Linux
  • AP-11649: XLS Writer throws IOException when writing to knime protocol and file does not exist.
  • AP-11581: H2O Integration: Ports of H2O scorer nodes are reversely ordered
  • AP-11574: Editing node annotations results in poor performance / UX
  • AP-11572: PCA Node Doesn't Always Reduce Dimensions Properly
  • AP-11532: Bad error message when Table Reader is unable to read .table file written with Parquet extension (but extension is not installed)
  • AP-11494: Tableau Preferences panel error with only Hyper nodes/sdk installed
  • AP-11485: Prediction Fusion: Improve description of minimum and maximum method
  • AP-11483: (Linux only) Entering annotation edit mode changes background of other view components in KNIME UI
  • AP-11477: SSH Connection Does Not Resolve Non-Existing Path Correctly
  • AP-11469: Rule Engine nodes - unclear description regarding comments
  • AP-11464: Missing Value node: Loading missing cell handlers causes ConcurrentModificationException
  • AP-11463: DL Keras Layer nodes allow white spaces in names but they do not work
  • AP-11462: Template link type cannot be updated from component
  • AP-11449: Options Use full path and Truncate prefix in Upload node not working correctly
  • AP-11443: Edit Nominal Domain doesn't show domain values in dialog on re-open
  • AP-11436: Container Input (Row) populates swagger example in wrong format
  • AP-11433: Port Objects are not copied in Streaming Mode
  • AP-11430: Unclear error message in the Container Input (Row) node
  • AP-11421: XGBoost: Fix dialog loading issue in learners
  • AP-11411: Renderer to Image: NPE on first dialog
  • AP-11402: H2O Integration: Rework Random Forest and Gradient Boosting Machine dialogs
  • AP-11381: Multiple Scroll bars in JavaScript Views
  • AP-11379: Excel Writer and Excel Sheet Appender automatically add xlsx, after xlsm
  • AP-11375: Renaming workflow with open report yields to lock on workflow
  • AP-11353: Interval Loop Start outputs rounding errors
  • AP-11327: Typo in "Category To Number" node description
  • AP-11276: Communication between Chromium browser window and Java sometimes 'hangs'
  • AP-11245: Source directory structure not deleted in KNIME Explorer when moving (and merging) directories
  • AP-11211: Typo in regexReplace function in String Manipulation and Column Expressions
  • AP-11208: DatabaseSession creation blocks workflow loading if db is not reachable
  • AP-11204: Potential race condition when canceling an executing node
  • AP-11201: JSON Path: table in configuration dialog not updating on macOS
  • AP-11196: Wait node ignores existing files
  • AP-11180: Extract Context Properties throws IlegalArgumentException in newly created workflows
  • AP-11178: Connections of MongoDB Nodes are not closed
  • AP-11170: JSON Reader: fails for UTF-8 encoded file with BOM
  • AP-11169: Corrupted display when creating a component
  • AP-11164: Java Snippet nodes: Expressions get unnecessarily re-compiled during configuration -- makes fast executing loops much slower (factor 10-100)
  • AP-11139: "Reference Column Resorter" fails when reference table contains unknown values
  • AP-10989: DL Python Network Creator can't create network with tf.string input/output
  • AP-10972: Jsonpath does not generate missing values for string columns
  • AP-10888: Gradient Boosted Trees to PMML produces invalid PMML
  • AP-10812: H2O Predictor (Classification) probability column names differ from other predictor nodes
  • AP-10532: DL TensorFlow: Reader: Tensor doesn't appear in ouputs drop-down of advanced settings
  • AP-10515: Text editor auto-completion is buggy on Windows, crashes on Mac
  • AP-10513: Stacked Area Chart does not respond to filters on excluded columns
  • AP-10471: DB SQL Executor accepts an empty SQL statement
  • AP-10411: Drag & drop files to KNIME Explorer stops working
  • AP-10199: Data Explorer nodes refers incorrectly to nominal columns charts as "histograms"
  • AP-7020: (Database) GroupBy node produces duplicate columns when doing Type Based aggregation
  • AP-6588: Problems with Linear Discriminant Analysis Node
  • AP-1590: NodeLogger isDebugEnabled always returns true
  • BD-921: (Big Data Extensions): Spark to Hive on HDP 3.1 fails when writing to non-default DB
  • BD-916: (Big Data Extensions): Parquet writer can be run with empty filename
  • BD-912: (Big Data Extensions): Parquet Reader nodes fail when executed in parallel
  • BD-898: (Big Data Extensions): Orc/Parquet reader fail to read from remote filesystems in streaming mode
  • BD-894: (Big Data Extensions): Parquet and ORC Reader fail to read from Kerberos-secured HDFS
  • BD-824: (Big Data Extensions): Parquet/ORC Reader/Writer Type mapping entry missing in node description
  • BD-812: (Big Data Extensions): ORC Reader should handle type conversion failures more gracefully
  • BD-810: (Big Data Extensions): Spark to Parquet/ORC doesn't exclude all columns in Partitions Tab if the spec changes
  • BD-180: (Big Data Extensions): HttpFS Connection node does not report on errors properly