Export

Saves a trained model and stores it as a work file in the workspace. Models can also be stored in TIBCO Streambase Artifact Management Server, if it is available and configured.

Information at a Glance

Category	Tools
Data source type	HD, DB
Sends output to other operators	No
Data processing tool	n/a

For detailed information about the types of model operators that you can export to the supported file types, see Model Export Formats.

Input

A trained model operator to export.

Restrictions

Currently, the following model configurations are not supported for PMML (Predictive Model Markup Language) export.

SVM Classification: Only Hadoop SVM is supported for linear kernel type models (not Gaussian or polynomial).
Alpine Forest: Only Hadoop Alpine Forest model export is currently supported.

Configuration

If an Artifact Management Server is not configured, users have the following parameters.

Parameter	Description
Notes	Any notes or helpful information about this operator's parameter settings. When you enter content in the Notes field, a yellow asterisk is displayed on the operator.
Export Format	Format of the exported model. Analytics Model Format (the default) PFA PMML
File Name	The file name for the exported model. Default value: @flow_name_model_0 For more information about the variables used for the default name, see Workflow Variables.
Ancestry Depth	A value between 1 (the default) and 3, representing how many (supported) operators prior to the model should be exported when using Analytics Model or PFA formats.

If an Artifact Management Server is available and configured, the following parameters are also available, in addition to those described above.

Export Location	Options: Workspace (the default) Streambase
Streambase User	The user name for your Streambase account. Default value: empty, but it can be configured by an administrator.
Streambase Password	The password for your Streambase account. Default value: empty, but can be configured by an administrator.
Overwrite?	Choose whether to overwrite an existing model in the specified Streambase project. If Yes (the default) and the Streambase Project has an existing model with the same name as specified in File Name, then AMS checks out the existing model and commits a new version. If No, the operator fails with a warning that a model already exists in the project.
Autocommit	Specify whether the change should be closed with a commit, or if it should leave the commit open for additional changes. Default value: true
Streambase Project	Select a Streambase project to store the exported model. The drop-down list includes all projects that are available to the user specified in the credentials above.

Output

Work file Output: The exported PMML (Predictive Model Markup Language), PFA (Portable Format for Analytics), or Analytics Model file is exported and saved as a work file within the current workspace, or in the specified project on the Artifact Management Server.

Click Download File to download the PMML file to the desktop.
Important: When you run the download file from within Safari browser, the generated PMML file is downloaded in XML format. Firefox or Chrome downloads the correct PMML format.
Visual Output: None.
Data Output: None. This is a terminal operator.