Cognite Function Examples

This repository contains a set of Jupyter Notebooks showcasing the use of Cognite Functions.

Requirements

The code examples in this repository are written in Jupyter notebooks. In order to display and run the code, you need to install jupyterlab and the Cognite Python SDK:

pip install jupyterlab cognite-sdk-experimental

Authentication

For all the examples here, we will be authenticating against CDF by using Azure Active Directory. Each notebook contains the relevant instructions.

Example 01: Create a basic Cognite Function from a Jupyter Notebook

In this example, we deploy a Cognite Function in the simplest possible fashion, by referring directly to a Python-function. This is sufficient for simple use cases where your function does not rely on additional dependencies.

Example 02: Create a Cognite Function from a folder.

In this example, we take it up a notch by deploying a function that is divided across multiple Python-files. We deploy an entire folder containing our source code. This method allows you to specify additional Python package dependencies if neccessary.

We also schedule this Cognite Function to run on a regular interval.

Example 03: Create a Cognite Function that performs entity matching on time-series and assets.

In this example, we look at something more akin to a real world use case. Here we deploy a Cognite Function that runs entity matching on some of our assets, mapping time-series to assets in CDF.

We also schedule this Cognite Function to run on a regular interval.