Creating a Dataset Template - Teradata Vantage

ClearScape Analytics™ ModelOps User Guide

Deployment
VantageCloud
VantageCore
Edition
Enterprise
IntelliFlex
VMware
Product
Teradata Vantage
Release Number
7.1
Published
December 2024
ft:locale
en-US
ft:lastEdition
2024-12-13
dita:mapPath
zdn1704469623418.ditamap
dita:ditavalPath
azq1671041405318.ditaval
dita:id
rgn1654191066978
lifecycle
latest
Product Category
ClearScape
Teradata recommends creating a default personal connection prior to creating a dataset template. See Adding a Connection.
  1. Open a project by selecting the project in the project list and select Datasets in the Navigation bar.
  2. Select Create Dataset Template.
  3. Select the default personal connection from the drop-down, then select Continue.
  4. In the Catalog tab, set the following properties:
    Property Description
    Name Dataset template name.
    Description Description of the dataset template.
    Feature catalog Catalog type to define a dataset template.

    Select Vantage for a Vantage type feature catalog and None for no catalog.

    The catalog type affects the wizard options.
    Tags Allows you to add tags with the dataset template for searches and filtering.
    Database Specify the database to store data features metadata.
    Table Specify the table to store data features metadata.
  5. Select Next.
  6. If Feature catalog is set to None:
    1. In the Config tab, set the following properties:
      Property Description
      Common Settings Custom properties: Specify custom properties in the form of key/value pair.
      Query template: Specify a query template that will fetch data for the dataset template.
      Scope Settings Specify properties for different scope levels: Train, Evaluate, Score.
      Custom properties: Specify custom properties in the form of key/value pair for each of the selected scope.
      Query template: Specify a query template that will fetch data to execute training, evaluation, and scoring jobs.
    2. Select Create to save the dataset template.
  7. If Feature catalog is set to Vantage, use the following steps 7 to 9 to continue creating a dataset template.
  8. In the Features tab, set the following properties:
    Property Description
    Query Define the SQL query to select features and entity key for catalog.
    Features Displays the metadata of the provided query and ability to perform further operations.
    1. You can define an SQL query to get an introspection on the features fetched directly from the database by selecting Run. Retrieved results displays in the following Features table with the ability to further specify collection by selecting or deselecting check boxes adjacent to each feature.
    2. A small key icon displays in parallel to one and only one row indicating it as the Entity Key. You can change entity key by switching between features using the same panel. An entity key cannot be a feature simultaneously, hence upon switching, a prompt appears to confirm selection.
    3. You can also retrieve actual result of features from SQL query defined earlier by selecting Preview data.
    4. Additionally, you can validate the features statistics with the ones stored in the database by selecting Validate statistics. For any feature missing statistics, a warning icon will appear on the row.
    5. Select Next to go to the next tab.
  9. In the Entity & target tab, set the following properties:
    Property Description
    Query Define the SQL query to select variables (entity sample and targets) for catalog.
    Features Displays the metadata of the provided query and ability to perform further operations.
    1. You can define an SQL query to retrieve desired variables by selecting Run. Retrieved results display in the Variables table with the ability to further specify sample by selecting or deselecting check boxes adjacent to each row. Entity Key selection and operations are exactly the same as Features step.
    2. You can also retrieve a sample of actual results from the combination of the SQL queries defined earlier, together with the resulting SQL queries, by selecting Preview data.
    3. Additionally, you can validate statistics for the current target selection by selecting Validate statistics. For any target missing statistics, a warning icon will appear on the row.
    4. Select Next to go to the next tab.
  10. In the Predictions tab, set the following properties:
    Property Description
    Database Name Define the database name to store predictions when using this dataset template.
    Table name Define the table name to store predictions when using this dataset template.
    Entity selection Specify an SQL query to load entity keys for scoring.
    1. You can retrieve actual result of sample from SQL query result with the entity keys for scoring defined earlier by selecting Preview data.
  11. Select Create.