Manage and find features

To learn more, run the "Example Feature Store workflow with sample data" Jupyter notebook in one of the following environments:

• Run the notebook in Colab.
• Run the notebook in a Vertex AI Workbench user-managed notebooks instance.
• View the notebook on GitHub.

Learn how to manage and find features.

Create a feature

Create a single feature for an existing entity type. To create multiple features in a single request, see Batch creating features.

Note: Vertex AI has a limit on the number features that you can create per entity type. For more information, see the Vertex AI Feature Store limits.

Web UI

1. In the Vertex AI section of the Google Cloud console, go to the Features page.

   Go to the Features page

2. Select a region from the Region drop-down list.
3. In the features table, view the Entity type column and click the entity type to add features to.
4. Click Add features to open the Add features pane.
5. Specify a name, value type, and (optionally) a description for the feature.
6. To enable feature value monitoring (Preview), under Feature monitoring, select Override entity type monitoring config and then enter the number of days between snapshots. This configuration overrides any existing or future monitoring configurations on the feature's entity type. For more information, see Feature value monitoring.
7. To add more features, click Add another feature.
8. Click Save.

REST & CMD LINE

To create a feature for an existing entity type, send a POST request by using the featurestores.entityTypes.features.create method.

Before using any of the request data, make the following replacements:

• LOCATION: Region where the featurestore is located, such as us-central1.
• PROJECT: Your project ID.
• FEATURESTORE_ID: ID of the featurestore.
• ENTITY_TYPE_ID: ID of the entity type.
• FEATURE_ID: An ID for the feature.
• DESCRIPTION: Description of the feature.
• VALUE_TYPE: The value type of the feature.

HTTP method and URL:

POST https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID?featureId=FEATURE_ID

Request JSON body:

{
  "description": "DESCRIPTION",
  "valueType": "VALUE_TYPE"
}

To send your request, choose one of these options:

curl

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Save the request body in a file called request.json, and execute the following command:

curl -X POST \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
-H "Content-Type: application/json; charset=utf-8" \
-d @request.json \
"https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID?featureId=FEATURE_ID"

PowerShell

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Save the request body in a file called request.json, and execute the following command:

$cred = gcloud auth application-default print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
  -Method POST `
  -Headers $headers `
  -ContentType: "application/json; charset=utf-8" `
  -InFile request.json `
  -Uri "https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID?featureId=FEATURE_ID" | Select-Object -Expand Content

You should see output similar to the following. You can use the OPERATION_ID in the response to get the status of the operation.

{
  "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/operations/OPERATION_ID",
  "metadata": {
    "@type": "type.googleapis.com/google.cloud.aiplatform.v1.CreateFeatureOperationMetadata",
    "genericMetadata": {
      "createTime": "2021-03-02T00:04:13.039166Z",
      "updateTime": "2021-03-02T00:04:13.039166Z"
    }
  }
}

Python

To learn how to install and use the client library for Vertex AI, see Vertex AI client libraries. For more information, see the Vertex AI Python API reference documentation.

View on GitHub Feedback

from google.cloud import aiplatform


def create_feature_sample(
    project: str,
    location: str,
    feature_id: str,
    value_type: str,
    entity_type_id: str,
    featurestore_id: str,
):

    aiplatform.init(project=project, location=location)

    my_feature = aiplatform.Feature.create(
        feature_id=feature_id,
        value_type=value_type,
        entity_type_name=entity_type_id,
        featurestore_id=featurestore_id,
    )

    my_feature.wait()

    return my_feature

Additional languages

You can install and use the following Vertex AI client libraries to call the Vertex AI API. Cloud Client Libraries provide an optimized developer experience by using each supported language's natural conventions and styles.

• Vertex AI SDK for Python
• Java
• Node.js

Batch create features

Create features in bulk for an existing type. For batch creation requests, Vertex AI Feature Store creates multiple features at once, which is faster for creating a large number of features compared to the featurestores.entityTypes.features.create method.

Note: Vertex AI has a limit on the number features that you can create per entity type. For more information, see the Vertex AI Feature Store limits.

Web UI

See creating a feature.

REST & CMD LINE

To create one or more features for an existing entity type, send a POST request by using the featurestores.entityTypes.features.batchCreate method, as shown in the following sample.

Before using any of the request data, make the following replacements:

• LOCATION: Region where the featurestore is located, such as us-central1.
• PROJECT: Your project ID.
• FEATURESTORE_ID: ID of the featurestore.
• ENTITY_TYPE_ID: ID of the entity type.
• PARENT: The resource name of the entity type to create the features under. Required format:
  projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID
• FEATURE_ID: An ID for the feature.
• DESCRIPTION: Description of the feature.
• VALUE_TYPE: The value type of the feature.
• DURATION: (Optional) The interval duration between snapshots in seconds. The value must end with an `s`.

HTTP method and URL:

POST https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features:batchCreate

Request JSON body:

{
  "requests": [
    {
      "parent" : "PARENT_1",
      "feature": {
        "description": "DESCRIPTION_1",
        "valueType": "VALUE_TYPE_1",
        "monitoringConfig": {
          "snapshotAnalysis": {
            "monitoringInterval": "DURATION"
          }
        }
      },
      "featureId": "FEATURE_ID_1"
    },
    {
      "parent" : "PARENT_2",
      "feature": {
        "description": "DESCRIPTION_2",
        "valueType": "VALUE_TYPE_2",
        "monitoringConfig": {
          "snapshotAnalysis": {
            "monitoringInterval": "DURATION"
          }
        }
      },
      "featureId": "FEATURE_ID_2"
    }
  ]
}

To send your request, choose one of these options:

curl

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Save the request body in a file called request.json, and execute the following command:

curl -X POST \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
-H "Content-Type: application/json; charset=utf-8" \
-d @request.json \
"https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features:batchCreate"

PowerShell

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Save the request body in a file called request.json, and execute the following command:

$cred = gcloud auth application-default print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
  -Method POST `
  -Headers $headers `
  -ContentType: "application/json; charset=utf-8" `
  -InFile request.json `
  -Uri "https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features:batchCreate" | Select-Object -Expand Content

You should see output similar to the following. You can use the OPERATION_ID in the response to get the status of the operation.

{
  "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/operations/OPERATION_ID",
  "metadata": {
    "@type": "type.googleapis.com/google.cloud.aiplatform.v1.BatchCreateFeaturesOperationMetadata",
    "genericMetadata": {
      "createTime": "2021-03-02T00:04:13.039166Z",
      "updateTime": "2021-03-02T00:04:13.039166Z"
    }
  }
}

Python

To learn how to install and use the client library for Vertex AI, see Vertex AI client libraries. For more information, see the Vertex AI Python API reference documentation.

View on GitHub Feedback

from google.cloud import aiplatform


def batch_create_features_sample(
    project: str,
    location: str,
    entity_type_id: str,
    featurestore_id: str,
    sync: bool = True,
):

    aiplatform.init(project=project, location=location)

    my_entity_type = aiplatform.featurestore.EntityType(
        entity_type_name=entity_type_id, featurestore_id=featurestore_id
    )

    FEATURE_CONFIGS = {
        "age": {"value_type": "INT64", "description": "User age"},
        "gender": {"value_type": "STRING", "description": "User gender"},
        "liked_genres": {
            "value_type": "STRING_ARRAY",
            "description": "An array of genres this user liked",
        },
    }

    my_entity_type.batch_create_features(feature_configs=FEATURE_CONFIGS, sync=sync)

Additional languages

You can install and use the following Vertex AI client libraries to call the Vertex AI API. Cloud Client Libraries provide an optimized developer experience by using each supported language's natural conventions and styles.

• Vertex AI SDK for Python
• Java
• Node.js

List features

List all features in a given location. To search for features across all entity types and featurestores in a given location, see the Searching for features method.

Web UI

1. In the Vertex AI section of the Google Cloud console, go to the Features page.

   Go to the Features page

2. Select a region from the Region drop-down list.
3. In the features table, view the Features column to see the features in your project for the selected region.

REST & CMD LINE

To list all features for a single entity type, send a GET request by using the featurestores.entityTypes.features.list method.

Before using any of the request data, make the following replacements:

• LOCATION: Region where the featurestore is located, such as us-central1.
• PROJECT: Your project ID.
• FEATURESTORE_ID: ID of the featurestore.
• ENTITY_TYPE_ID: ID of the entity type.

HTTP method and URL:

GET https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features

To send your request, choose one of these options:

curl

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

curl -X GET \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
"https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features"

PowerShell

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

$cred = gcloud auth application-default print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
  -Method GET `
  -Headers $headers `
  -Uri "https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features" | Select-Object -Expand Content

You should receive a JSON response similar to the following:

{
  "features": [
    {
      "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID_1",
      "description": "DESCRIPTION",
      "valueType": "VALUE_TYPE",
      "createTime": "2021-03-01T22:41:20.626644Z",
      "updateTime": "2021-03-01T22:41:20.626644Z",
      "labels": {
        "environment": "testing"
      },
      "etag": "AMEw9yP0qJeLao6P3fl9cKEGY4ie5-SanQaiN7c_Ca4QOa0u7AxwO6i75Vbp0Cr51MSf"
    },
    {
      "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID_2",
      "description": "DESCRIPTION",
      "valueType": "VALUE_TYPE",
      "createTime": "2021-02-25T01:27:00.544230Z",
      "updateTime": "2021-02-25T01:27:00.544230Z",
      "labels": {
        "environment": "testing"
      },
      "etag": "AMEw9yMdrLZ7Waty0ane-DkHq4kcsIVC-piqJq7n6A_Y-BjNzPY4rNlokDHNyUqC7edw"
    },
    {
      "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID_3",
      "description": "DESCRIPTION",
      "valueType": "VALUE_TYPE",
      "createTime": "2021-03-01T22:41:20.628493Z",
      "updateTime": "2021-03-01T22:41:20.628493Z",
      "labels": {
        "environment": "testing"
      },
      "etag": "AMEw9yM-sAkv-u-jzkUOToaAVovK7GKbrubd9DbmAonik-ojTWG8-hfSRYt6jHKRTQ35"
    }
  ]
}

Additional languages

You can install and use the following Vertex AI client libraries to call the Vertex AI API. Cloud Client Libraries provide an optimized developer experience by using each supported language's natural conventions and styles.

• Vertex AI SDK for Python
• Java
• Node.js

Search for features

Search for features based on one or more of their properties, such as feature ID, entity type ID, or feature description. Vertex AI Feature Store searches across all featurestores and entity types in a given location. You can also limit results by filtering on specific featurestores, value types, and labels.

To list all features, see Listing features.

Web UI

1. In the Vertex AI section of the Google Cloud console, go to the Features page.

   Go to the Features page

2. Select a region from the Region drop-down list.
3. Click the features table's Filter field.
4. Select a property to filter on such as Feature, which returns features that contain a matching string anywhere in their ID.
5. Type a value for the filter and then press enter. Vertex AI Feature Store returns results in the features table.
6. To add additional filters, click the Filter field again.

REST & CMD LINE

To search for features, send a GET request by using the featurestores.searchFeatures method. The following sample uses multiple search parameters, written as featureId:test AND valueType=STRING. The query returns features that contain test in their ID and whose values are of type STRING.

Before using any of the request data, make the following replacements:

• LOCATION: Region where the featurestore is located, such as us-central1.
• PROJECT: Your project ID.

HTTP method and URL:

GET https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores:searchFeatures?query="featureId:test%20AND%20valueType=STRING"

To send your request, choose one of these options:

curl

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

curl -X GET \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
"https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores:searchFeatures?query="featureId:test%20AND%20valueType=STRING""

PowerShell

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

$cred = gcloud auth application-default print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
  -Method GET `
  -Headers $headers `
  -Uri "https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores:searchFeatures?query="featureId:test%20AND%20valueType=STRING"" | Select-Object -Expand Content

You should receive a JSON response similar to the following:

{
  "features": [
    {
      "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/featurestore_demo/entityTypes/testing/features/test1",
      "description": "featurestore test1",
      "createTime": "2021-02-26T18:16:09.528185Z",
      "updateTime": "2021-02-26T18:16:09.528185Z",
      "labels": {
        "environment": "testing"
      }
    }
  ]
}

Additional languages

You can install and use the following Vertex AI client libraries to call the Vertex AI API. Cloud Client Libraries provide an optimized developer experience by using each supported language's natural conventions and styles.

• Vertex AI SDK for Python
• Java
• Node.js

View feature details

View details about a feature, such as its value type or description. If you use the Google Cloud console and have feature monitoring enabled, you can also view the distribution of feature values over time.

Web UI

1. In the Vertex AI section of the Google Cloud console, go to the Features page.

   Go to the Features page

2. Select a region from the Region drop-down list.
3. In the features table, view the Features column to find the feature that you want to view details for.
4. Click the name of a feature to view its details.
5. To view its metrics, click Metrics. Vertex AI Feature Store provides feature distribution metrics for the feature.

REST & CMD LINE

To get details about a feature, send a GET request by using the featurestores.entityTypes.features.get method.

Before using any of the request data, make the following replacements:

• LOCATION: Region where the featurestore is located, such as us-central1.
• PROJECT: Your project ID.
• FEATURESTORE_ID: ID of the featurestore.
• ENTITY_TYPE_ID: ID of the entity type.
• FEATURE_ID: ID of the feature.

HTTP method and URL:

GET https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID

To send your request, choose one of these options:

curl

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

curl -X GET \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
"https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID"

PowerShell

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

$cred = gcloud auth application-default print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
  -Method GET `
  -Headers $headers `
  -Uri "https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID" | Select-Object -Expand Content

You should receive a JSON response similar to the following:

{
  "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID",
  "description": "DESCRIPTION",
  "valueType": "VALUE_TYPE",
  "createTime": "2021-03-01T22:41:20.628493Z",
  "updateTime": "2021-03-01T22:41:20.628493Z",
  "labels": {
    "environment": "testing"
  },
  "etag": "AMEw9yOZbdYKHTyjV22ziZR1vUX3nWOi0o2XU3-OADahSdfZ8Apklk_qPruhF-o1dOSD",
  "monitoringConfig": {}
}

Additional languages

You can install and use the following Vertex AI client libraries to call the Vertex AI API. Cloud Client Libraries provide an optimized developer experience by using each supported language's natural conventions and styles.

• Vertex AI SDK for Python
• Java
• Node.js

Delete a feature

Delete a feature and all of its values.

Web UI

1. In the Vertex AI section of the Google Cloud console, go to the Features page.

   Go to the Features page

2. Select a region from the Region drop-down list.
3. In the features table, view the Feature column and find the feature to delete.
4. Click the name of the feature.
5. From the action bar, click Delete.
6. Click Confirm to delete the feature and its values.

REST & CMD LINE

To delete a feature, send a DELETE request by using the featurestores.entityTypes.features.delete method.

Before using any of the request data, make the following replacements:

• LOCATION: Region where the featurestore is located, such as us-central1.
• PROJECT: Your project ID.
• FEATURESTORE_ID: ID of the featurestore.
• ENTITY_TYPE_ID: ID of the entity type.
• FEATURE_ID: ID of the feature.

HTTP method and URL:

DELETE https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID

To send your request, choose one of these options:

curl

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

curl -X DELETE \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
"https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID"

PowerShell

Note: Ensure you have set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the path for your service account private key file.

Execute the following command:

$cred = gcloud auth application-default print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
  -Method DELETE `
  -Headers $headers `
  -Uri "https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT/locations/LOCATION/featurestores/FEATURESTORE_ID/entityTypes/ENTITY_TYPE_ID/features/FEATURE_ID" | Select-Object -Expand Content

You should receive a JSON response similar to the following:

{
  "name": "projects/PROJECT_NUMBER/locations/LOCATION/featurestores/FEATURESTORE_ID/operations/OPERATION_ID",
  "metadata": {
    "@type": "type.googleapis.com/google.cloud.aiplatform.v1.DeleteOperationMetadata",
    "genericMetadata": {
      "createTime": "2021-02-26T17:32:56.008325Z",
      "updateTime": "2021-02-26T17:32:56.008325Z"
    }
  },
  "done": true,
  "response": {
    "@type": "type.googleapis.com/google.protobuf.Empty"
  }
}

Additional languages

You can install and use the following Vertex AI client libraries to call the Vertex AI API. Cloud Client Libraries provide an optimized developer experience by using each supported language's natural conventions and styles.

• Vertex AI SDK for Python
• Java
• Node.js

What's next

• Learn how to batch ingest feature values.
• Learn how to monitor ingested feature values over time.
• Learn how to serve features through online serving or batch serving.
• Troubleshoot common Vertex AI Feature Store issues.