SearchAsync()()
Perform a vector similarity search.
Invocation
await collection.SearchAsync<T>(vectorFieldName, vectors, metricType, limit, parameters = null, cancellationToken = default);
Parameters
| Parameter | Description | Type | Required |
|---|---|---|---|
vectorFieldName | The name of the vector field to search in. | string | True |
vectors | The set of vectors to send as input for the similarity search. | IReadOnlyList<ReadOnlyMemory<T>> | True |
metricType | Method used to measure the distance between vectors during search. Must correspond to the metric type specified when building the index. | SimilarityMetricType | True |
limit | The maximum number of records to return, also known as 'topk'. Must be between 1 and 16384. | int | True |
parameters | Various additional optional parameters to configure the similarity search. | SearchParameters | False |
cancellationToken | The token to monitor for cancellation requests. The default value is CancellationToken.None. | CancellationToken | False |
Return
The results of the vector similarity search.
Example
var results = await milvusClient.GetCollection("book").SearchAsync(
vectorFieldName: "book_intro",
vectors: new ReadOnlyMemory<float>[] { new[] { 0.1f, 0.2f } },
SimilarityMetricType.L2,
// the sum of `offset` in `parameters` and `limit` should be less than 16384.
limit: 10,
new SearchParameters
{
OutputFields = { "title" },
ConsistencyLevel = ConsistencyLevel.Strong,
Offset = 5,
ExtraParameters = { ["nprobe"] = "1024" }
});