milvus-logo
LFAI
< Docs
  • Python

construct_from_dict()

This operation constructs a CollectionSchema object from a dictionary representation.

Request Syntax

construct_from_dict(
    raw: dict
)

PARAMETERS:

  • raw (dict)

    A dictionary containing the raw data to construct the collection schema.

RETURN TYPE:

CollectionSchema

RETURNS:

A CollectionSchema object.

EXCEPTIONS:

  • MilvusException

    This exception will be raised when any error occurs during this operation.

Examples

from pymilvus import DataType, FieldSchema, CollectionSchema

# Define fields and create a schema
primary_key = FieldSchema(
    name="id",
    dtype=DataType.INT64,
    is_primary=True,
)

vector = FieldSchema(
    name="vector",
    dtype=DataType.FLOAT_VECTOR,
    dim=768,
)

# Create dictionary representation 
schema_dict = {
    "fields": [     
        primary_key.to_dict(),
        vector.to_dict()                
    ]
}  

# Reconstruct schema from dictionary 
schema = CollectionSchema.construct_from_dict(schema_dict)  
# schema is now a CollectionSchema instance reconstructed from the dictionary 
print(schema)

# Output
# {'auto_id': False, 'description': '', 'fields': [{'name': 'id', 'description': '', 'type': <DataType.INT64: 5>, 'is_primary': True, 'auto_id': False}, {'name': 'vector', 'description': '', 'type': <DataType.FLOAT_VECTOR: 101>, 'params': {'dim': 768}}]}