Overview
Dask is a flexible and high-performance Python library for parallel computing. It scales and accelerates big data processing with other Python-based data science libraries, such as Pandas, Numpy, and Scikit-Learn.
This instructor-led, live training (online or onsite) is aimed at data scientists and software engineers who wish to use Dask with the Python ecosystem to build, scale, and analyze large datasets.
By the end of this training, participants will be able to:
- Set up the environment to start building big data processing with Dask and Python.
- Explore the features, libraries, tools, and APIs available in Dask.
- Understand how Dask accelerates parallel computing in Python.
- Learn how to scale the Python ecosystem (Numpy, SciPy, and Pandas) using Dask.
- Optimize the Dask environment to maintain high performance in handling large datasets.
Format of the Course
- Interactive lecture and discussion.
- Lots of exercises and practice.
- Hands-on implementation in a live-lab environment.
Course Customization Options
- To request a customized training for this course, please contact us to arrange.
Requirements
- Experience with data analysis
- Python programming experience
Audience
- Data scientists
- Software engineers
Course Outline
Introduction
- Overview of Dask features and advantages
- Parallel computing in Python
Getting Started
- Installing Dask
- Dask libraries, components, and APIs
- Best practices and tips
Scaling NumPy, SciPy, and Pandas
- Dask arrays examples and use cases
- Chunks and blocked algorithms
- Overlapping computations
- SciPy stats and LinearOperator
- Numpy slicing and assignment
- DataFrames and Pandas
Dask Internals and Graphical UI
- Supported interfaces
- Scheduler and diagnostics
- Analyzing performance
- Graph computation
Optimizing and Deploying Dask
- Setting up adaptive deployments
- Connecting to remote data
- Debugging parallel programs
- Deploying Dask clusters
- Working with GPUs
- Deploying Dask on cloud environments
Troubleshooting
Summary and Next Steps