hopsworks.core.superset_api #

SupersetApi #

API for managing Superset datasets, charts, and dashboards.

This class provides methods to create, read, update, and delete Superset resources directly from within Hopsworks. Authentication is handled automatically using the project user's Superset credentials stored in Hopsworks secrets.

Example

import hopsworks
import json

project = hopsworks.login()
superset = project.get_superset_api()

# 1. List existing datasets in Superset
datasets = superset.list_datasets()

# 2. Create a dataset backed by an existing database table
dataset = superset.create_dataset(
    database_id=1,                       # ID of the Superset database connection
    table_name="transactions_1",         # table in the online feature store
    schema="myproject_onlinedb",         # schema (project's online DB name)
)
dataset_id = dataset["id"]

# 3. Create a chart that visualises the dataset
chart = superset.create_chart(
    slice_name="Transactions over time",
    viz_type="echarts_timeseries_line",  # Superset viz type
    datasource_id=dataset_id,
    params=json.dumps({                  # chart config as a JSON string
        "metrics": [{"label": "count", "expressionType": "SQL", "sqlExpression": "COUNT(*)"}],
        "groupby": [],
        "time_column": "event_time",
        "time_grain_sqla": "P1D",
        "row_limit": 1000,
    }),
)
chart_id = chart["id"]

# 4. Create a dashboard and add the chart to it
dashboard = superset.create_dashboard(
    dashboard_title="Feature Monitoring",
    published=True,
)
dashboard_id = dashboard["id"]

# Add the chart to the dashboard by updating the chart
superset.update_chart(chart_id, dashboards=[dashboard_id])

# 5. Clean up
superset.delete_chart(chart_id)
superset.delete_dataset(dataset_id)
superset.delete_dashboard(dashboard_id)

Returned by

Project.get_superset_api

list_databases #

list_databases() -> dict

List all Superset database connections visible to the current user.

RETURNS	DESCRIPTION
`dict`	A dict containing the list of databases under the "result" key.

create_dataset #

create_dataset(
    database_id: int,
    table_name: str,
    schema: str | None = None,
    sql: str | None = None,
    description: str | None = None,
    owners: list[int] | None = None,
) -> dict

Create a Superset dataset.

PARAMETER	DESCRIPTION
`database_id`	ID of the Superset database connection. TYPE: `int`
`table_name`	Name of the table (or virtual dataset). TYPE: `str`
`schema`	Database schema name. TYPE: `str \| None` DEFAULT: `None`
`sql`	SQL expression for virtual datasets. TYPE: `str \| None` DEFAULT: `None`
`description`	Optional description. TYPE: `str \| None` DEFAULT: `None`
`owners`	Optional list of Superset user IDs as owners. TYPE: `list[int] \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`dict`	The created dataset as a dict (Superset API response).

get_dataset #

get_dataset(dataset_id: int) -> dict

Get a Superset dataset by ID.

PARAMETER	DESCRIPTION
`dataset_id`	The dataset ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	The dataset as a dict.

list_datasets #

list_datasets() -> dict

List all Superset datasets visible to the current user.

RETURNS	DESCRIPTION
`dict`	A dict containing the list of datasets under the "result" key.

update_dataset #

update_dataset(dataset_id: int, **kwargs: Any) -> dict

Update a Superset dataset.

Pass fields to update as keyword arguments (e.g., description, sql, table_name, owners).

PARAMETER	DESCRIPTION
`dataset_id`	The dataset ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	The updated dataset as a dict.

delete_dataset #

delete_dataset(dataset_id: int) -> dict

Delete a Superset dataset.

PARAMETER	DESCRIPTION
`dataset_id`	The dataset ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	Empty dict on success.

create_chart #

create_chart(
    slice_name: str,
    viz_type: str,
    datasource_id: int,
    params: str,
    datasource_type: str = "table",
    description: str | None = None,
    dashboards: list[int] | None = None,
    owners: list[int] | None = None,
) -> dict

Create a Superset chart.

PARAMETER	DESCRIPTION
`slice_name`	Display name of the chart. TYPE: `str`
`viz_type`	Visualization type (e.g., "table", "big_number", "echarts_timeseries"). TYPE: `str`
`datasource_id`	ID of the dataset to use as data source. TYPE: `int`
`params`	JSON string of chart configuration parameters. TYPE: `str`
`datasource_type`	Type of datasource (default "table"). TYPE: `str` DEFAULT: `'table'`
`description`	Optional description. TYPE: `str \| None` DEFAULT: `None`
`dashboards`	Optional list of dashboard IDs to add this chart to. TYPE: `list[int] \| None` DEFAULT: `None`
`owners`	Optional list of Superset user IDs as owners. TYPE: `list[int] \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`dict`	The created chart as a dict.

get_chart #

get_chart(chart_id: int) -> dict

Get a Superset chart by ID.

PARAMETER	DESCRIPTION
`chart_id`	The chart ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	The chart as a dict.

list_charts #

list_charts() -> dict

List all Superset charts visible to the current user.

RETURNS	DESCRIPTION
`dict`	A dict containing the list of charts under the "result" key.

update_chart #

update_chart(chart_id: int, **kwargs: Any) -> dict

Update a Superset chart.

Pass fields to update as keyword arguments (e.g., slice_name, viz_type, params, owners).

PARAMETER	DESCRIPTION
`chart_id`	The chart ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	The updated chart as a dict.

delete_chart #

delete_chart(chart_id: int) -> dict

Delete a Superset chart.

PARAMETER	DESCRIPTION
`chart_id`	The chart ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	Empty dict on success.

create_dashboard #

create_dashboard(
    dashboard_title: str,
    published: bool = False,
    slug: str | None = None,
    position_json: str | None = None,
    json_metadata: str | None = None,
    css: str | None = None,
    owners: list[int] | None = None,
) -> dict

Create a Superset dashboard.

PARAMETER	DESCRIPTION
`dashboard_title`	Display title of the dashboard. TYPE: `str`
`published`	Whether the dashboard is published. TYPE: `bool` DEFAULT: `False`
`slug`	Optional URL slug. TYPE: `str \| None` DEFAULT: `None`
`position_json`	Optional JSON string defining the layout. TYPE: `str \| None` DEFAULT: `None`
`json_metadata`	Optional JSON string with dashboard metadata. TYPE: `str \| None` DEFAULT: `None`
`css`	Optional custom CSS. TYPE: `str \| None` DEFAULT: `None`
`owners`	Optional list of Superset user IDs as owners. TYPE: `list[int] \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`dict`	The created dashboard as a dict.

get_dashboard #

get_dashboard(dashboard_id: int) -> dict

Get a Superset dashboard by ID.

PARAMETER	DESCRIPTION
`dashboard_id`	The dashboard ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	The dashboard as a dict.

list_dashboards #

list_dashboards() -> dict

List all Superset dashboards visible to the current user.

RETURNS	DESCRIPTION
`dict`	A dict containing the list of dashboards under the "result" key.

update_dashboard #

update_dashboard(dashboard_id: int, **kwargs: Any) -> dict

Update a Superset dashboard.

Pass fields to update as keyword arguments (e.g., dashboard_title, published, position_json, json_metadata, css, owners).

PARAMETER	DESCRIPTION
`dashboard_id`	The dashboard ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	The updated dashboard as a dict.

delete_dashboard #

delete_dashboard(dashboard_id: int) -> dict

Delete a Superset dashboard.

PARAMETER	DESCRIPTION
`dashboard_id`	The dashboard ID. TYPE: `int`

RETURNS	DESCRIPTION
`dict`	Empty dict on success.