Vector Retrieval Service:Search for vectors

Use the DashVector console to search for vectors

1. Log on to the DashVector console.

2. In the left-side navigation pane, click Clusters. On the Clusters page, find the collection that you want to manage and click Details in the Operation column.

3. In the left-side bar, click Similar Vector Search. Then, configure parameters and click Search. The following table describes the parameters.

   Parameter	Parameter name in API	Description	Required
   Query Vector	vector	The vector. Example: [1.00,2.00,3.00,4.00]. Note The number of dimensions and data type of the vector must be the same as those of the current collection.	Yes
   Filter conditions	filter	The conditional filter, which must comply with the syntax of the SQL WHERE clause. For more information, see Conditional filtering.	No
   Partition	partition	The partition name. You can select a partition based on your business requirements. Default value: default.	Yes
   TopK	topk	The maximum number of vectors that can be returned. Valid values: 1 to 1024. Default value: 10.	Yes

   The following table describes the parameters in the search results.

Parameter	Parameter name in API	Description
Distance	score	The similarity of vectors. The value description of this parameter varies based on the distance measurement method. For more information, see What is a vector? The returned results are sorted in descending order by similarity.
Primary Key ID	id	The primary key ID of the similar vector.
Vector	vector	The vector. Example: [1.00,2.00,3.00,4.00].
Attribute	fields	The attribute information in the JSON format. Example: {"price":100,"type":"dress"}.
Partition	partition	The partition to which the similar vector belongs.

Use SDKs to search vectors

For more information about how to search for vectors by using the SDK for Python, see Search for documents.

For more information about how to search for vectors by using the SDK for Java, see Search for documents.

Use the HTTP API to search vectors

For more information, see Search for documents.