Table of Contents
The Workflow component of the AdvancedMiner System is a graphical environment for data analysis, modelling and processing, designed to accelerate the process of creating and testing analytical models. Workflow's user interface uses graphical nodes to represent various steps and elements of the analytical process, such as data sources, data processing, modelling and visualization. These tasks are completed without need to write scripts.
Before starting to work with a workflow document a user needs to make sure the following objects are already defined in the application:
Aliases are displayed in the application window Services . To create a new alias click node Aliases and from the context menu choose . You can create a Gdbase database - the one which comes with AdvancedMiner. To create local database give it a name in the field Alias Name, choose a proper database driver in the field Database Driver and a directory where the database files should be stored - Database URL.
Projects and repositories are displayed in the the Window Projects. To create a project from the application menu and choose For Categories choose Projects and for Projectschoose Basic Project. In the field Project Name put the name of your project.
To create a repository, from the application menu choose In the field Repository Name put the name of your repository and in the field Local Directory give directory where the repository objects will be sored.
Workflow documents are stored in the file format in the project directory of the operating system. In AdvancedMiner Workflow documents are created within projects. To create a Workflow file click the node of the selected project and from its context menu choose . Next choose Categories Other and File Type Workflow. Next give the name for the Workflow document and choose its alias and repository. Workflow documents are listed in a project window under the node of the project where they were created.
Using Workflow module users create analytical schemas that process and visualise data. The schemas consist of nodes and directed connections between them. The schemas are created by selecting nodes from a palette window and dragging them on the work area. Connections are also created by using the drag and drop technique. Start dragging a connection from the anchor sign from the left side of the node.
While using the module, users should remember the following:To check the current repository and databse alias, click the work area and from the context menu choose .
Any node in the work area can be checked for proper configuration and connections with other nodes. To check the node, from its context menu choose .
To check which nodes can be after a certain node, from the context menu choose item
To run processing according to the workflow schema, click the last node for which processing should be executed and click the run button or press F6. To process the whole schema, mark all the nodes (Ctrl+A) and run processing.
The processed nodes are marked with a green tick. By default they will not be calculated again unless user changed their settings or settings of the preceding nodes. However one can force the system to run processing again. To do it from the node's context menu choose .
Nodes for which processing resulted in errors or warnings are marked with a red exclamation mark. This mark can also indicate problems with node's configuration. Information about errors or warnings can be obtained by clicking this exclamation mark. Application logs processing events. Log entries are given in the Log window. In case of problems with processing, user should analyze entries in the Log window and Output window.
To parameterize the node choose or items from the node's context menu.
Connections between the nodes have their statuses. They can be active, restricted or inactive. To check or change status of the connection double-click it. If the status is active (default) then calculations will be performed starting from from the node which was not yet processed. In case of inactive status the node from which a connection starts will not be calculated as well all the following nodes. In case of the restricted connection, a node from which the connection starts will be executed if before execution user clicks a node after this connection. Otherwise the connection works like restricted connection.
In case of big schema, use Navigator (Ctrl+7) window to move the work area to the different place of the schema.
If the schema contains many parallel paths, then the processing can be executed in parallel. Maximum number of parallel processing threads is given in the properties of the Workflow schema available from the context menu after right-clicking the empty place of the work area.