Matrix size controls what is allowed
A matrix operation begins with dimensions. Addition and subtraction require matrices of the same size. Multiplication requires the columns of the first matrix to match the rows of the second matrix. If the sizes do not fit, the operation is not defined no matter how neat the entries look.
Write the dimensions before calculating: rows by columns. A 2 by 3 matrix and a 3 by 2 matrix contain the same number of entries, but they are not the same shape.
Addition and subtraction are entry-by-entry
Matrix addition and subtraction combine matching positions. The top-left entry combines with the top-left entry, the next entry with the next entry, and so on. This is why the matrices need the same dimensions.
Matrix multiplication is not entry-by-entry
Matrix multiplication uses rows from the first matrix and columns from the second matrix. The order matters. AB may be defined while BA is not, and even when both are defined, they may not be equal. That makes matrix multiplication different from ordinary number multiplication.
When a multiplication result looks unexpected, check the order before checking every arithmetic entry. Many matrix mistakes are order mistakes.
Transpose flips rows and columns
A transpose turns rows into columns and columns into rows. It does not change the values themselves, only their positions. Transpose is useful when aligning dimensions or rewriting data orientation.
Scalar multiplication changes every entry
A scalar is a single number multiplied by every matrix entry. If the scalar is -2, every entry changes sign and doubles in size. This operation does not have the same dimension restrictions as multiplying two matrices.
Determinants belong to square matrices
A determinant is defined for square matrices. It can tell whether a matrix is invertible and appears in geometry, systems of equations, and transformations. A non-square matrix does not have a determinant in the ordinary sense.
Inverse matrices undo matrix multiplication
An inverse matrix works like a reciprocal for compatible matrix multiplication. Not every square matrix has an inverse. If the determinant is zero, the inverse does not exist. The calculator result should be checked for that warning before the inverse is used in a solution.
RREF is a solving form
Reduced row echelon form rewrites a matrix through row operations. It is often used to solve systems of linear equations or understand rank. The entries in the final form should be read as a transformed version of the original matrix, not a separate dataset.
If the matrix represents equations, keep the variable order and augmented column clear. Swapping columns can change the meaning of the final row-reduced form.
Rank describes independent information
Matrix rank tells how many independent rows or columns are present. Low rank can signal repeated information or dependent equations. This matters in systems, transformations, and data problems where not every row adds new information.
Eigenvalue work needs careful interpretation
Eigenvalues and eigenvectors describe directions that a matrix transformation stretches or scales without changing direction. They are usually studied after basic matrix operations. If the page shows them for a 2 by 2 matrix, keep the original matrix visible because small entry changes can shift the result.
For most routine matrix work, confirm the basic operation first: dimensions, order, and copied entries. Advanced outputs are only useful after those fundamentals are correct.