Mathematics
Linear Transformations and Their Geometric Interpretations
Quick fact
A linear transformation is completely determined by what it does to the basis vectors—just knowing where the unit vectors go tells you where every vector goes. This is why a matrix captures the transformation in a compact grid of numbers.
Why this is interesting
Imagine stretching, rotating, or squishing a photo on your phone—these are linear transformations. But what makes a transformation 'linear', and why do they always keep straight lines straight?
Read the full explanation
Understanding Linear Transformations and Their Geometric Interpretations
Think of a grid drawn on a rubber sheet. A linear transformation is like stretching or rotating that sheet, but with strict rules: the origin (0,0) must stay fixed, and all parallel lines must remain parallel. More formally, a function T from one vector space to another is linear if it satisfies two rules: adding vectors first then transforming gives the same result as transforming each and then adding (T(u+v) = T(u) + T(v)), and scaling a vector by a number then transforming is the same as transforming then scaling (T(c·v) = c·T(v)). These two rules are what preserve the grid's structure. Because of these rules, you only need to know where the basis vectors (like (1,0) and (0,1) in 2D) go. Each transformation is represented by a matrix whose columns are the images of the basis vectors. For example, a 90° rotation sends (1,0) to (0,1) and (0,1) to (-1,0), so the matrix is [[0,-1],[1,0]]. Multiply any vector by this matrix, and you get the rotated vector.
A deeper explanation
The real power of linear transformations lies in their geometric effects and how matrices encode them. Every matrix defines a linear transformation, and conversely, every linear transformation between finite-dimensional spaces can be represented by a matrix. The determinant of the matrix tells you how area (or volume) is scaled: if |det| 1, the transformation expands space; if |det| < 1, it shrinks; if det = 0, it flattens space onto a lower dimension. The eigenvectors of the transformation are the directions that stay on the same line (only scaled), which explains why they are so important in understanding the transformation's action. Geometrically, linear transformations can be visualized as combinations of basic operations: scaling, rotation, shearing, reflection, and projection. For instance, a shear keeps one axis fixed but shifts the other axis proportionally, preserving area (det = 1). Projection collapses one dimension, so det = 0. This geometric view makes it clear why the set of all linear transformations into the same space is closed under composition—multiplying matrices corresponds to applying one transformation after another. This understanding underpins everything from computer graphics (rotations, translations via affine matrices) to data science (principal component analysis uses linear transformations to rotate data).