Stride

Q: How deeply does Fast Python: High performance techniques for large datasets cover stride?

Defines strides as byte steps and demonstrates them with reshaping, transposition, slicing, and image views.

As explained in Fast Python: High performance techniques for large datasets

A stride is the step interval between successive positions or elements accessed, selected, or processed along a dimension or sequence.

A NumPy stride is the number of bytes to step along a dimension to reach the next element. Together, shape and strides describe how an array view presents data in memory. section 4.4 section 4.1.2

Reading stride values

A one-dimensional array of 4-byte integers has a stride of 4. Reshaping the same data into a 2 × 5 matrix gives it a stride for each dimension; transposing the matrix changes those values while preserving shared memory. section 4.1.2

Stride changes can describe different views of existing array data:

Reverse an array or image dimension by using a negative stride. section 4.1.3
Use slicing, transposition, or rotation to present array data differently without copying it in the demonstrated cases. section 4.1.3
Treat numpy.lib.stride_tricks.as_strided with extreme care because arbitrary shape and stride values can be wrong. section 4.1.3

How Fast Python: High performance techniques for large datasets covers it

Cover of Fast Python: High performance techniques for large datasets

liveBook manning.com

Read Fast Python: High performance techniques for large datasets on liveBook Get the book at manning.com

Also explained in

liveBook manning.com

Math and Architectures of Deep LearningThe step size with which a convolutional kernel moves across the input, affecting the output size in convolution and transposed convolution.
liveBook manning.com

Learn Generative AI with PyTorchThe step-size parameter that controls how far a convolutional operation moves and, in transposed convolution, how much upsampling occurs.
liveBook manning.com

Deep Learning with PyTorch, Second EditionTensor metadata describing its shape and how elements are laid out and indexed in storage, including step size between positions.
liveBook manning.com

Grokking AI Algorithms, Second EditionA stride is the spatial step size by which a convolutional filter moves across an input.
liveBook manning.com

Build a Large Language Model (From Scratch)The step size by which the sliding window advances between sampled training examples.
liveBook manning.com

Natural Language Processing in Action, Second EditionA stride is the number of input positions by which a convolution kernel moves between successive applications.
liveBook manning.com

Python How-To

Compare how 8 books explain stride →

Related concepts

tensor shapecontrasts with · A tensor shape is an ordered sequence of dimension sizes specifying the extent of a tensor or array along each axis.
offsetproduces · An offset is a position or displacement measured relative to a reference point, such as within an ordered sequence, memory region, or layout.
dilated convolutioncontrasts with · A dilated convolution applies a filter whose sampling positions are spaced apart, expanding the receptive field without proportionally increasing kernel size or parameters.
boundary checkused by · A boundary check is a test that verifies whether a value or operation lies within an allowed range or valid data extent.

This explainer was generated by Manning's knowledge graph from the text of Fast Python: High performance techniques for large datasets. The section links point to the parts of the book it draws on.