Dataset properties

You can use Dataset properties to document the unique qualities, features, or characteristics of a dataset. These properties can be added as a dictionary when logging the dataset or after it's created.

Properties are optional and work with Dataset.origin, Dataset.clean and Dataset.modeling by assigning to the properties parameter, as shown below:

origin_resource = FileResource(paths="items.csv")
properties = {"test":10, "QA":"Approved", "Bool":True}
# Assigning properties while logging an origin dataset 
origin_dataset = Dataset.origin(
                  name = "Origin_Dataset",
                  resource = origin_resource,


You can also update or add dataset properties after a dataset has been logged. You must re-log your dataset to an iteration to save your updated dataset properties.

# Updating or adding properties after logging origin dataset = {"test":1, "QA":"Not started", "Bool":False}

# Re-log dataset to save updated properties