insert()

This method inserts entities into a collection. You can insert data through either a list of rows or a list of columns.

Invocation

insert(
    collection_name,
    data,
    batch_size,
    progress_bar,
    timeout
)

Parameters

Parameter	Description	Type	Required
collection_name	Name of the collection to which data is inserted.	String	True
data	A list of dictionaries to insert. Each dictionary represents an entity. All fields defined in the schema are required and should be included as keys in the dictionaries. You can also add dynamic keys for any fields that are not defined in the schema.	list[Dictionary]	True
batch_size	Size of entities to insert in a batch.	Integer	False
progress_bar	Whether to display a progress bar.	Boolean	False
timeout	An optional duration of time in seconds to allow for the RPC. If it is set to None, the client keeps waiting until the server responds or error occurs.	Float	False

Return

A list of primary keys inserted.

Raises

DataNotMatchException: Error if the data has misssing fields.

MilvusException: Error if a general Milvus error occurred.

Example

• Insert data through a list of rows:

  # Prepare data as a list of rows
  import json
  
  with open('/path/to/downloaded/dataset') as f:
      data = json.load(f)
      list_of_rows = data['rows'][0]
  
  # Create a MilvusClient instance
  from pymilvus import MilvusClient
  
  client = MilvusClient(uri, token)
  
  client.insert(list_of_rows)
  

• Insert data through a list of columns:

  # Prepare data as a list of columns
  import json
  
  with open('/path/to/downloaded/dataset') as f:
      data = json.load(f)
      keys = data['rows'][0].keys()
      entity = data['rows'][0]
      list_of_columns = [ [entity.get(x)] for x in keys]
  
  # Create a MilvusClient instance
  from pymilvus import MilvusClient
  
  client = MilvusClient(uri, token)
  
  client.insert(list_of_columns)