Search()

This method conducts a vector similarity search.

Invocation

client.Search(ctx, collName, partNames, expr, outputFields, vectors, fieldName, metricType, topK, sp)

Parameters

Parameter	Description	Type
ctx	Context to control API invocation process	context.Context
collName	Name of the collection to search in	String
partNames	List of names of the partitions to search on. All partition will be searched if it is left empty.	Slice of string
expr	Boolean expression to filter the data	String
outputFields	List of names of fields to output	Slice of string
vectors	Vectors to search with	Slice of entity.Vector
fieldName	Name of the vector field to search on	String
metricType	Metric type to calculate distance with	entity.MetricType
topK	Number of nearest records to return	INT
sp	Specific search parameter(s) of the index on the vector field	entity.SearchParam

Return

A slice of SearchResult that contains the search result, one record per vector.

Errors

err: error in the process (if any). Possible errors are listed below:

• ErrClientNotReady, error that the client is not connected.

• ErrCollectionNotExists, error that collection with the specified name does not exist.

• error that API invocation failed.

Example

searchResult, err := milvusClient.Search(
    context.Background(),                    // ctx
    "book",                                  // CollectionName
    []string{},                              // partitionNames
    "",                                      // expr
    []string{"book_id"},                     // outputFields
    []entity.Vector{entity.FloatVector([]float32{0.1, 0.2})}, // vectors
    "book_intro",                            // vectorField
    entity.L2,                               // metricType
    2,                                       // topK
    sp,                                      // sp
)
if err != nil {
    log.Fatal("fail to search collection:", err.Error())
}
fmt.Printf("%#v\n", searchResult)
for _, sr := range searchResult {
    fmt.Println(sr.IDs)
    fmt.Println(sr.Scores)
}