Datasets

Containers of the dataset type can store any kind of object inside. You can put any object type inside, including datasets and dataframes.

Creating datasets

You can create such a container as follows:

[Copy Code](javascript:void(0)) Python
from research_sdk import Dataset, DataStorageInterface, DataStorageType with DataStorageInterface.create(DataStorageType.Datacat) as storage: with Dataset(name="Test dataset", storage=storage) as ds: # I assume you already have some existing object to add in data_object variable ds.add(data_object)

A dataset has the following functions to work with:

• add
• get
• ennumerate

Adding an object to a dataset

The add function adds an object and flushes it to datacatalog immediately.

[Copy Code](javascript:void(0)) Python
def add(self, obj: DataObject) -> DataObjectInfo

The returned DataObjectInfo contains the information about added object.

[Copy Code](javascript:void(0)) Python
DataObjectType = Type["DataObject"] @dataclass class DataObjectInfo: id: str name: str type: DataObjectType

where

• id is the identifier of the added object.
• name is the object name.
• type is the python type of the object.

Getting an object from the dataset

[Copy Code](javascript:void(0)) Python
def get(self, name: str) -> Optional["DataObject"]:

Not implemented.

Enumerating objects in a dataset

[Copy Code](javascript:void(0)) Python
def enumerate(self) -> List["DataObjectInfo"]:

Not implemented.