Accessing data for custom data science models
After Running training and scoring jobs that calculate values for output attributes, you can do one of the following to access the data.
- Create an export job and select the relevant data object to export. You can then review the values in the output attributes.
- Sample data for the model outputs can be viewed using the Data viewer page. You can view a maximum of 200 records from a data object.
To verify and review with an export job:
- Follow the steps for Creating export jobs.
- When you get to Step 2: Data payload, for Payload type, select Data object.
- In the field for Data object, select the relevant data object that stores output attributes.
When the data object is exported, review the attributes that store output values for the model.
To verify and review using the Data viewer page:
-
Click the Oracle icon
in the bottom-right corner to open the navigation menu.
- Select Data viewer.
- Use the drop-down menu for Select or search for data object and select the relevant data object that stores output attributes.
- Click Filters
. The Filters dialog displays. - Use the drop-down menu to select or search for the attributes that store output values.
- Click Add
. - Use the drop-down menu for the output variable attribute and select the Is not empty operator.
- Click Apply.
- Click Column settings
. - Click the checkbox for attributes that store output values.
- Click Apply.
You can review values for up to 200 records.