NumPy
NumPy is an open-source Python library for numerical computing, providing multidimensional arrays and operations on them.
Fast Python: High performance techniques for large datasets
- How it is used
- Provides efficient arrays for Python data analysis and a starting point before custom code re-implementation.
- How deep it goes
- Explains views, broadcasting, configuration, and threading for performance, while assuming familiarity and deferring a comprehensive introduction.
- Examples
- Uses image manipulation and pandas fare calculations to illustrate array-based processing.
- What you can do after
- Apply it via NumPy to use array operations and views, inspect numerical-library configuration, and assess threading trade-offs.
Read how Fast Python: High performance techniques for large datasets explains NumPy →
Statistics Every Programmer Needs
NumPy is a Python library for efficient numerical computation using multidimensional arrays and mathematical operations.
- How it is used
- Supports Python-based numerical computation within quantitative analysis and predictive modeling workflows.
- How deep it goes
- Demonstrates arithmetic, matrix multiplication, and repeated transition calculations, without detailing NumPy internals.
- Examples
- Uses an initial probability vector, a transition matrix, matrix squaring, and US cellular market-share forecasts.
- What you can do after
- Apply it via Python to automate arithmetic and matrix calculations for multi-period state-probability forecasts.
Pandas Workout: 200 exercises to make you a stronger data analyst
NumPy is a Python library providing multidimensional arrays and numerical operations.
- How it is used
- Provides reproducible numerical arrays for constructing and analyzing pandas Series.
- How deep it goes
- Demonstrates seeded integer generation and array shape, without developing broader NumPy array operations.
- Examples
- Uses ten seeded integers, month labels, and a Series average as the concrete exercise.
- What you can do after
- Apply NumPy's random generator to create bounded, reproducible integer arrays.
Read how Pandas Workout: 200 exercises to make you a stronger data analyst explains NumPy →
Geometry for Programmers
NumPy is a Python library for numerical computing with multidimensional arrays and mathematical operations.
- How it is used
- Supports efficient numeric computation, matrix algebra, interpolation workflows, and array-based geometric calculations.
- How deep it goes
- Demonstrates matrix solving, coordinate grids, and vectorization, while stopping short of detailed numerical-method internals.
- Examples
- Uses interpolating data, matrices, linear systems, circle signed distance functions, linspace, meshgrid, and vectorized branching.
- What you can do after
- Apply NumPy to construct arrays, solve linear systems, evaluate functions over grids, and process scalar logic across arrays.
Deep Learning with JAX
A Python library for array-based numerical computing.
- How it is used
- Supports Python numerical computing and practical tensor-processing pipelines such as image loading, filtering, and saving.
- How deep it goes
- Explains the array API and image-array properties through a worked pipeline, while deferring several NumPy–JAX differences.
- Examples
- Uses denoising an image, loading a photo, inspecting shape and rank, and comparing NumPy code with jax.numpy.
- What you can do after
- Apply it via NumPy arrays to represent images, inspect their structure, and build basic numerical-processing pipelines.
Build Your Own Robot: Using Python, CRICKIT, and Raspberry Pi
NumPy is a Python library for numerical computing that provides multidimensional arrays and operations on numerical data.
- How it is used
- Supplies image-array data for OpenCV camera capture and drawing tasks.
- How deep it goes
- Demonstrates ndarray type, shape, dtype, and zero-filled image creation without broader NumPy operations.
- Examples
- Uses a 480×640×3 captured frame and a black array canvas for circles, lines, rectangles, and text.
- What you can do after
- Apply it via OpenCV to create and inspect image arrays for camera and drawing workflows.
Read how Build Your Own Robot: Using Python, CRICKIT, and Raspberry Pi explains NumPy →
Python Concurrency with asyncio
NumPy is a Python library for numerical computing with multidimensional arrays, matrices, and vectorized mathematical operations.
- How it is used
- Uses NumPy for efficient numerical array operations and for selectively accelerating CPU-bound workloads with threads.
- How deep it goes
- Explains arrays, axes, reshaping, performance, GIL release, profiling, and benchmarking, but defers detailed implementation guarantees.
- Examples
- Demonstrates row means over a reshaped four-billion-point array and distributes row calculations through a thread pool.
- What you can do after
- Apply it via NumPy array operations, then evaluate threaded alternatives with profiling and benchmarks.
CUDA for LLMs
NumPy is a Python library for numerical computing with multidimensional arrays.
- How it is used
- Uses NumPy to make neural-network training mathematics explicit and bridge high-level autograd to low-level implementation.
- How deep it goes
- Implements the forward pass, activations, cross-entropy loss, and full backpropagation.
- Examples
- Compares NumPy and PyTorch training by loss reduction and final accuracy despite different random initializations.
- What you can do after
- Evaluate a manual backpropagation implementation against PyTorch using training loss and final accuracy.
Also covered in
- Deep Learning with R, Third Edition
- Build a Robo-Advisor with Python (From Scratch)
- The Complete Obsolete Guide to Generative AI
- Python How-To
- The Well-Grounded Data Analyst
- The Quick Python Book, Fourth Edition
- Grokking Machine Learning, Second EditionIn early access (MEAP)
- Grokking GPU ProgrammingIn early access (MEAP)
What to know first
- arrayAn array is an ordered collection of elements organized in one or more dimensions and accessible by position or index.
- PythonPython is a high-level, general-purpose, dynamically typed programming language with readable syntax and support for multiple programming paradigms.
- CC is a general-purpose procedural programming language that provides low-level control over memory and system operations.
- automatic differentiationAutomatic differentiation is a computational technique that evaluates derivatives of a program by systematically applying the chain rule to its elementary operations.
Related concepts
- C++uses · C++ is a statically typed, general-purpose programming language that supports procedural, object-oriented, generic, and low-level systems programming.
- Pythonused by · Python is a high-level, general-purpose, dynamically typed programming language with readable syntax and support for multiple programming paradigms.
- pandasused by · pandas is an open-source Python library providing labeled data structures and tools for manipulating and analyzing tabular and other structured data.
- arrayuses · An array is an ordered collection of elements organized in one or more dimensions and accessible by position or index.
- vectorizationuses · Vectorization is the transformation of data or computations into vector form so they can be processed collectively by vector-oriented operations.
- Cuses · C is a general-purpose procedural programming language that provides low-level control over memory and system operations.
Questions about NumPy
What is NumPy?
NumPy is an open-source Python library for numerical computing, providing multidimensional arrays and operations on them.
Which Manning books explain NumPy?
Fast Python: High performance techniques for large datasets, Statistics Every Programmer Needs, Pandas Workout: 200 exercises to make you a stronger data analyst, Geometry for Programmers, Deep Learning with JAX, Build Your Own Robot: Using Python, CRICKIT, and Raspberry Pi, Python Concurrency with asyncio and CUDA for LLMs.
What should I know before learning NumPy?
Manning books that teach NumPy usually introduce array, Python, C and automatic differentiation first.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.







