Dask Tutorial

This repository contains an introduction to Dask and tutorials to use Dask arrays and stackstac to retrieve a large number of satellite scenes from a STAC API using Dask. This is part of the course on Advanced Geospatial Analytics with Python taught since Fall 2023 at Clark University.

Requirements

You need to have Docker installed on your machine.

Instructions

It's recommended to pull the Docker image from Dockerhub. Otherwise, if you prefer, you can build your own image using the instructions in the following section.

docker pull hamedalemo/dask-tutorial:1.3

docker run -it -p 8888:8888 -p 8787:8787 hamedalemo/dask-tutorial:1.3

Port 8787 is used by Dask Dashboard.

• Copy the Jupyter Lab url and paste it in your browser.
• Open dask_intro.ipynb, stackstac.ipynb or dask_dataframe.ipynb and follow the instructions.

Build the Docker image:

Windows users: It is highly recommended that you pull the Docker image, there seems to be an issue with conda-forge on WSL.

docker build -t dask-tutorial .

Run the container as following after switching to the repository's directory locally:

docker run -it -p 8888:8888 -p 8787:8787 dask-tutorial

• Copy the Jupyter Lab url and paste it in your browser.
• Open dask_intro.ipynb, stackstac.ipynb or dask_dataframe.ipynb and follow the instructions.