Skip to content

Quickstart

Quickstart is a collection of modules that configure Cognite Data Fusion with data pipelines to ingest and contextualize data and set up industrial tools to make the data ready for applications.

Demo

:::caution Alpha

The quickstart demo is currently in alpha and is likely to change. Do not use this in production systems.

:::

The quickstart package has a Notebook interface to run it as a demo and in the CDF Notebook environment.

To run the quickstart package in the CDF notebook environment:

  1. Go to Admin -> My organization.
  2. Create a CDF service account. Store the Client ID.
  3. Click the newly created service account and create new credentials. Store the Client Secret.
  4. Go to Data management -> Build solutions and select Jupyter Notebook.
  5. Create a new Notebook.
  6. Copy-paste the content below into a cell and execute it.

Installing cognite-toolkit in the CDF notebook environment: 

import micropip

micropip.uninstall("prompt-toolkit")

%pip install cognite-toolkit
We uninstall prompt-toolkit because the Pyodide environment has an issue resolving dependencies. The dependency is reinstalled in a compatible version when you install cognite-toolkit.

Import and instantiate the Cognite ToolkitDemo: 

from cognite_toolkit.demo import CogniteToolkitDemo

toolkit = CogniteToolkitDemo()

Set up the configuration and credentials to run the quickstart package: 

# This is used in the extension of the Cognite Process Indusdry Model
my_organization = "<your_organization_name>"
client_id = "<your_client_id>" # The client ID of the service account
client_secret = "<your_client_secret>" # The client secret of the service account

Running the quickstart package: 

toolkit.quickstart(my_organization, client_id, client_secret)

The last cell does the equivalent of running cdf auth verify, cdf modules init (selecting quickstart and choosing example data), cdf build and cdf deploy.