What's New in Rulex 4.2

Import from R and RBridge Tasks

NEW

New Python Import Bridge and Python Bridge tasks have been introduced in Rulex 4.2 to facilitate communication with Python scripts.

The R Import Bridge and R Bridge tasks previously present in Rulex 3.2 have also been reintroduced, but with an improved performance of the external language connectors compared to the previous version.

Import from Cloudera Impala, Apache Spark, Teradata and Gupta SQLBase DB

NEW

Cloudera Impala, Apache Spark, Teradata and Gupta SQLBase DB have been added to the list of possible database distributions which can be used in database import and export operations.

Workflow Versioning

NEW

In previous versions of Rulex you could either duplicate your process every time you wanted to fix a version, or export the process with its date and time as a suffix. To avoid the human errors inherent in manual activities, and to minimize memory occupation, processes can now be versioned directly from the Rulex interface.

This first release of Rulex versioning allows you to save, update, merge and retrieve specific versions of a process, tag specific versions and revert to previous versions. The versioning system is based on the Git version control system.

Search AssistantNEW

A search feature is now available in Rulex, which allows you to search for a particular word or process variable within different processes. The search function shows all  occurrences of the searched word in the options, histories and task definitions. Changes can also be applied to all occurrences from a central point.

Another frequent problem that can occur is when you want to delete a process variable, but are unable to do so because it is in use. Thanks to the search function you can check which tasks use it and decide whether it can effectively be removed, and how.

Conditional Import from DB

NEW

A new task, called Conditional Import from DB, has been introduced, which performs an import from a database conditioned by a set of constraints defined through a dataset (each row is a constraint) or a set of rules in if-then form.

Workflow execution parameters - priority policy

NEW

Two ways of managing priorities have been added to the Process Execution Parameter pane. 

These two options help solve ambiguous situations, where either the parent relationship or priority level are considered more important.

See example below.

Model to dataset new conversion tasks

NEW

Two new tasks have been added to Rulex to make it possible import and export models to file and database:

• The Convert Model to Dataset task produces a dataset that contains all the information included in the model, which can then be visualized in a Data Manager or exported to an external item, such as a file or database table.

• The Convert Dataset to Model task takes a dataset with a proper format and outputs the corresponding model, which can then be used in an Apply Model task in order to derive responses for specific samples.

New ODBC library to deal with DB

CHANGE

The library used to connect Rulex to the working and external database has been changed in the kernel from SQLAPI++ to ODBC.

The ODBC library represents the world standard as regards DB connections, used by most commercial and open source software for DB access.

Revision of all the http/cloud remote connectors

CHANGE

All the http/cloud connectors (http/ftp/sharepoint/s3/hdfs) have been rewritten to increase their stability and speed. For example, the SharePoint connector is now twice as fast and the S3 connector three times faster than in Rulex 4.1.

Bulk import from excel, text file, json via remote connections

CHANGE

The bulk import feature has been extended to include import operations performed from remote connections.

A new dialog box has been added from which you can select the multiple files and folders from which you want to import data.

Usability improvements made to some Rulex tasks

CHANGE

The following tasks have been modified to improve usability:

• K-Nearest Neighbor

• K-Nearest Neighbor Regression

• LLM One-Class

• Classification SVM

• Regression SVM