Linear Algebra 11 | Matrices [dark version]

Matrices enter linear algebra as a practical way to organize many numbers so they can later solve systems of linear equations. The core idea is simple: a matrix is a two-dimensional table of real numbers (and, later in the series, complex numbers too). Each entry sits at an intersection of a row and a column, indexed as a_{i j}, where i identifies the row and j identifies the column. That row/column indexing matters because it determines how matrix operations act entry-by-entry.

A matrix is described by its shape: m rows by n columns. That means it contains m·n positions, each filled with a real number. The transcript emphasizes that matrices need not be square; they can be rectangular. For example, with m = 2 and n = 3, there are six entries arranged in two rows and three columns. In general, the bottom-right entry carries the index a_{m n}, reinforcing the row/column structure.

All matrices of a fixed shape form a set, denoted using the notation R^{m×n}. Here, the superscript indicates the number of rows and columns that define the rectangle layout. Once that set is fixed, the next step is defining operations that make matrices behave like vectors under algebraic rules.

Two operations are introduced: matrix addition and scalar multiplication. Addition works only for matrices of the same shape. Given two m×n matrices A and B, their sum C is another m×n matrix whose entries are computed by ordinary real-number addition in each position: c_{i j} = a_{i j} + b_{i j}. A quick numerical example with 2×2 matrices shows how each corresponding entry adds to produce the resulting matrix.

Scalar multiplication also preserves shape. If λ is a real number and A is an m×n matrix, then λA is the m×n matrix whose entries are λ times each entry of A: (λA)_{i j} = λ·a_{i j}. Again, the operation is defined entry-by-entry using standard real-number multiplication.

With these two operations in place, the transcript connects matrices to vector space structure. The set R^{m×n}, equipped with addition and scalar multiplication, satisfies the usual vector space axioms. Under addition, matrices form an abelian group: there is a zero matrix (the additive identity) and each matrix has an additive inverse obtained by negating every entry. Scalar multiplication is compatible with multiplying scalars in either order, and distributive laws hold—both for distributing over matrix addition and for distributing over scalar addition. Because these properties mirror the familiar behavior of vectors, matrices can be manipulated with the same kind of algebraic consistency.

The motivation then turns toward applications: matrices are introduced not as an abstract object, but as the tool that will later make systems of linear equations solvable. The next installment is set to use these matrix operations to tackle linear equations.