Dataset

class Dataset

Dataset on Redivis are the entity where data is stored. Datasets are made up of tables, non-tabular files, and various metadata. Datasets can be owned by a user or organization, and are version controlled.

Constructors

Organization.dataset(dataset_reference)	Construct a new dataset instance that references a dataset owned by an organization.
Organization.list_datasets([max_results])	Returns a list of Datasets owned by an organization.
User.dataset(dataset_reference)	Construct a new dataset instance that references a dataset owned by a user.
User.list_datasets([max_results])	Returns a list of Datasets owned by a user.

Examples

Basics

dataset = redivis.organization("Demo").dataset("US Fires")

# Will throw an error if the dataset doesn't exists
# Can first call dataset.exists() to check for existence
dataset.get()

print(dataset.properties)

Create

dataset = redivis.user("my_username").dataset("My dataset")

dataset.create(public_access_level="overview")

print(dataset.properties)

New version

dataset = redivis.user("my_username").dataset("My dataset")

dataset = dataset.create_next_version()

# We can upload new data to existing tables once we have a "next" version
with open("data.csv", "rb") as file:
    dataset.table("My table").upload("data.csv").upload_file(file)
    
dataset.release()

Query

dataset = redivis.organization("Demo").dataset("CMS 2014 Medicare Data")

# The home_health_agencies table is assumed to be within the dataset,
#   since it isn't otherwise qualified
query = dataset.query("""
    SELECT * FROM home_health_agencies
    WHERE state = 'CA'
""")

print(query.to_dataframe())

Attributes

organization	A reference to the Organization instance that constructed this dataset. Will be None if the dataset belongs to a user.
properties	A dict containing the API resource representation of the dataset. This will be fully populated after calling get(), create_next_version(), and release(), otherwise will be None. This will also be partially populated for datasets returned via the Organization.list_datasets and User.list_datasets methods
qualified_reference	The fully qualified reference for the dataset, which can be used in SQL queries or the REST API. E.g., demo.ghcn_daily_weather_data:v1_1:7br5
scoped_reference	The canonical reference for the dataset, without the username qualifier. E.g.,: ghcn_daily_weather_data:v1_1:7br5
user	A reference to the User instance that constructed this dataset. Will be None if the dataset belongs to an organization.

Methods

Dataset.create([*, public_access_level, ...])	Create a new dataset.
Dataset.create_next_version([*, if_not_exists])	Create a "next" (unreleased) version on the dataset. Data can only be uploaded to unreleased versions.
Dataset.delete()	Delete the dataset.
Dataset.exists()	Check whether the dataset exists.
Dataset.get()	Get the dataset, populating the properties attribute on the current instance.
Dataset.list_tables([max_results])	List all tables in the dataset.
Dataset.query(query_string)	Create a query scoped to the dataset.
Dataset.release()	Release the next version of the dataset.
Dataset.table(table_reference)	Create a reference to a specific table within the dataset.
Dataset.update([*, name, public_access_level, ...])	Update certain attributes on the dataset.