The following summarizes the most noteworthy improvements of KNIME v2.7, for a complete list of enhancements and bug fixes please have a look at the changelog.
Note that in light of the update in our Eclipse version, a clean installation of KNIME will be required. An update of KNIME 2.6 is not possible.
Core
Java KNIME now runs on Java 7 for Windows and Linux systems (Mac stays on Java 6) |
Eclipse Eclipse update 3.7 increases stability on Mac and some Linux systems. |
BIRT BIRT 3.7 brings Open Office support among other new features |
GUI
KNIME Explorer View as default Workflow Repository The “KNIME Explorer” View is now the default Workflow Repository. It is still possible to call back the old “Workflow Projects” panel via the “View” item in the top menu. | |
Placeholder Node for Missing Nodes Missing nodes in a workflow are now represented by a placeholder node, briefly explaining which node is missing. | |
Database Nodes
New Nodes “Database Delete” and “Database Update” The “Database Delete” and “Database Update” node respectively deletes and updates the database table records matching the input data. | |
New Supported Data Types Database nodes now support data columns of Boolean and Blob type as well. |
|
JFreeChart Nodes
More Plot Options in JFreeChart Nodes JFreeChart nodes have now more setting options in the “General Plot Options” tab of their configuration window. For example, it is now possible to set:
| |
“Line Chart" Node with Color support
The “Line Chart” node now has two input ports:
| |
R Integration
New Nodes In R-> Local there are a number of new nodes to import:
| |
R Version 2.15.2 The R Binaries from the R Package Extension install R version 2.15.2. | |
KNIME Labs
Special Cases in File Handling KNIME Labs has now a category named “File Handling” including file handling utilities, like:
| |
New Nodes: Tree Ensemble (Regression) Learn/Predict based off of an ensemble of regression trees. Learner output includes:
| |
Text Mining and Network Analytics “Text Processing” and “Network Mining” categories in KNIME Labs have also been enriched with a number of new nodes. |
Similarity Search
New Node: Similarity Search Use bit vectors, numeric descriptors or predefined distance matricies to query a reference table for it’s most similar (or disparate) entries. | |
Reporting
“Toggle Breadcrumb” button to show report item hierarchy In BIRT 3.7 the “Toggle Breadcrumb” button is now available. This button activates a bar on top of the report layout that shows the hierarchy position for the selected report item. | |
Quickform Nodes
New Node “Multi-selection QuickForm” This QuickForm node:
| |
New Node “TextArea Output” The “TextArea Output” node builds a formatted text, including flow variable values, to be output at the end of the workflow execution on the web portal. | |
New Node “Image Output” Similarly to the “TextArea Output” node, the “Image Outout” node takes an image as input data and outputs the same scaled image at the end of the workflow execution on the web portal. Configuration Settings require the image scaling parameters. | |
New “Hide” Option in QuickForm Nodes Most QuickForm nodes have now a new option “Hide in Wizard”. This option, when enabled, hides the QuickForm node’s mask during the server wizard execution and uses the default selection values. | |
New Layout in “String Radio Buttons” Node The mask layout of the “String Radio Buttons” node now displays the options vertically. | |
Enterprise
New nodes: Variable, Text, and Image Output Nodes In the KNIME Webportal, the new Quickform Output nodes enable the display of intermediate results in the form of an image, variable, or text in the KNIME Webportal. | |
File Support on Server Support for files located on the KNIME server. You can now upload arbitrary data files to the KNIME Server. | |
Execution of Multiple Workflows The commercial KNIME report batch executor allows now for the execution of multiple workflows and reports at once. The run configuration is defined by an xml file that lists the workflows to run and their parameters. Workflows can also be executed in parallel. |