Skip to main content

ColumnFloatVectorArray

A ColumnFloatVectorArray instance represents an ArrayOfVector float-vector column whose rows contain float-vector values with a shared dimension. Nullable semantics are enforced in v3.0.0: nullability flags and valid-value bookkeeping are real operations instead of stubs.

go
type ColumnFloatVectorArray struct {
}

BUILDER METHODS:

  • AppendValue(value any) error

    This appends one row supplied as []entity.FloatVector or [][]float32.

  • IsNull(idx int) (bool, error)

    This reports whether the row at the given index is null.

  • Nullable() bool

    This reports whether the column accepts null values.

  • SetNullable(nullable bool)

    This sets whether the column accepts null values.

  • ValidCount() int

    This returns the number of non-null rows.

  • ValidateNullable() error

    This verifies the internal nullable bookkeeping and returns an error on mismatch.

METHODS:

  • Slice(start, end int) Column

Example​

Demonstrates ColumnFloatVectorArray usage.

go
import (
"fmt"

"github.com/milvus-io/milvus/client/v3/column"
)

values := column.NewColumnFloatVectorArray("embeddings", 4, [][][]float32{{{0.1, 0.2, 0.3, 0.4}}})
fmt.Println(values.Len())
Minimum SDK versionv3.0.x
Was this page helpful?