Launching a Service¶
Once you create a dataset and define transformations in the workspace, you can use the Predictive Learning notebook functionality and launch an external service. Introductory notebooks are provided with initial instructions to get you started. Refer to the documentation for the service for further instructions.
Predictive Learning supports the following services:
- Zeppelin Notebook
- Tensorboard
- Spark History
- Jupyter Notebook
How to Launch a Service¶
Follow these steps to launch a service:
- Navigate to the Manage Analytics Workspace page. The Cluster Configuration page appears.
- Select a Cluster Type and click Start Cluster. When the cluster is running, the Cluster Status changes to RUNNING and the Launch a Service field and Launch button display.
- Open a saved workspace, with transformations and columns selected, and select a service from the Services drop-down list.
- Click Launch. An introductory page for the service you selected displays with instructions on getting started.
- Refer to the Help provided for the selected service for more information.
For the Apache Zeppelin Introduction Notebook¶
Once the Notebook has launched, click the Introduction Notebook for information on how to use and access the Notebook in the Predictive Learning environment.
Once launched, you can write your own script for performing an analysis. You can also create multiple notebooks, as needed.
PrL Introduction Notebook Illustration¶
This is an example of the Predictive Learning Introduction Notebook:
Jupyter Notebook¶
When selecting Jupyter Notebook in the Service drop down, a new browser tab will open, providing access to your previously saved notebooks.
Except where otherwise noted, content on this site is licensed under the Development License Agreement.