Oreilly – Fast Python, Video Edition 2023-11

Fast Python, Video Edition. This course provides a set of advanced techniques for improving the speed and efficiency of Python code, especially in the areas of big data and machine learning. In this course, you will learn how to write optimized Python code, effectively use the NumPy and pandas libraries, rewrite critical parts of your code with Cython, design persistent data structures, and tune your code for different architectures. You will also learn how to implement GPU-based computing in Python. This course is designed for experienced professionals and includes practical examples such as rewriting games with Cython and building a MapReduce framework from scratch. By learning these concepts, you will be able to fix inefficient code, take advantage of multithreading capabilities, and optimize datasets without sacrificing accuracy. Finally, this course shows you how using modern hardware can transform old inefficient patterns and provide more efficient solutions. These skills are essential for entering the world of machine learning and large-scale data analytics, helping you perform better with fewer resources. About the course: Fast Python is a collection of techniques for speeding up Python, with an emphasis on Big Data applications. Following clear examples and detailed instructions, you’ll learn how to use common libraries like NumPy and pandas in more efficient ways and transform data for efficient storage and I/O. More importantly, Fast Python takes a holistic approach to performance, so you’ll see how to optimize the entire system, from code to architecture.

What you will learn:

• Writing efficient pure Python code
• Optimize NumPy and pandas libraries for better performance
• Rewrite critical code using Cython to increase speed
• Designing stable data structures for efficient data management
• Adapting Python code for different hardware architectures
• Implementing GPU Computing in Python to Accelerate Machine Learning Applications

Who is this course suitable for?

• Intermediate Python programmers who are familiar with the basics of concurrency.
• People looking to improve the performance of their Python code in Big Data projects.
• Machine Learning professionals who want to optimize the execution of their complex applications.
• Developers interested in leveraging the power of GPUs for Python computing.

Fast Python, Video Edition Course Specifications

• Publisher: Oreilly
• Instructor: Tiago Antao
• Training level: Beginner to advanced
• Training duration: 8 hours and 52 minutes

Course headings

• Part 1. Foundational Approaches
  Chapter 1. An urgent need for efficiency in data processing
  Chapter 1. Modern computing architectures and high-performance computing
  Chapter 1. Working with Python’s limitations
  Chapter 1. A summary of the solutions
  Chapter 1. Summary
• Chapter 2. Extracting maximum performance from built-in features
  Chapter 2. Profiling code to detect performance bottlenecks
  Chapter 2. Optimizing basic data structures for speed: Lists, sets, and dictionaries
  Chapter 2. Finding excessive memory allocation
  Chapter 2. Using laziness and generators for big-data pipelining
  Chapter 2. Summary
• Chapter 3. Concurrency, parallelism, and asynchronous processing
  Chapter 3. Implementing a basic MapReduce engine
  Chapter 3. Implementing a concurrent version of a MapReduce engine
  Chapter 3. Using multiprocessing to implement MapReduce
  Chapter 3. Tying it all together: An asynchronous multithreaded and multiprocessing MapReduce server
  Chapter 3. Summary
• Chapter 4. High-performance NumPy
  Chapter 4. Using array programming
  Chapter 4. Tuning NumPy’s internal architecture for performance
  Chapter 4. Summary
• Part 2. Hardware
  Chapter 5. Re-implementing critical code with Cython
  Chapter 5. A whirlwind tour of Cython
  Chapter 5. Profiling Cython code
  Chapter 5. Optimizing array access with Cython memory views
  Chapter 5. Writing NumPy generalized universal functions in Cython
  Chapter 5. Advanced array access in Cython
  Chapter 5. Parallelism with Cython
  Chapter 5. Summary
• Chapter 6. Memory hierarchy, storage, and networking
  Chapter 6. Efficient data storage with Blosc
  Chapter 6. Accelerating NumPy with NumExpr
  Chapter 6. The performance implications of using the local network
  Chapter 6. Summary
• Part 3. Applications and Libraries for Modern Data Processing
  Chapter 7. High-performance pandas and Apache Arrow
  Chapter 7. Techniques to increase data analysis speed
  Chapter 7. pandas on top of NumPy, Cython, and NumExpr
  Chapter 7. Reading data into pandas with Arrow
  Chapter 7. Using Arrow interop to delegate work to more efficient languages ​​and systems
  Chapter 7. Summary
• Chapter 8. Storing big data
  Chapter 8. Parquet: An efficient format to store columnar data
  Chapter 8. 8. Dealing with larger-than-memory datasets the old-fashioned way
  Chapter 8. Zarr for large-array persistence
  Chapter 8. Summary
• Part 4. Advanced Topics
  Chapter 9. Data analysis using GPU computing
  Chapter 9. Using Numba to generate GPU code
  Chapter 9. Performance analysis of GPU code: The case of a CuPy application
  Chapter 9. Summary
• Chapter 10. Analyzing big data with Dask
  Chapter 10. The computational cost of Dask operations
  Chapter 10. Using Dask’s distributed scheduler
  Chapter 10. Summary
• Appendix A. Setting up the environment
• Appendix A. Installing your own Python distribution
• Appendix A. Using Docker
• Appendix A. Hardware considerations
• Appendix B. Using Numba to generate efficient low-level code
• Appendix B. Writing explicitly parallel functions in Numba
• Appendix B. Writing NumPy-aware code in Numba

Course images

Sample course video

Download link

Download file – 829 MB

File(s) password: www.downloadly.ir

File size

829 MB