Square matrix
n columns and n rows
Diagonal matrix
square matrix with non-zero elements only on main diagonal
Band matrix
Non-zero elements only on main diagonal and few neighboring diagonals
Triangular matrix
square matrix with non-zero elements only below or above the main diagonal , main diagonal included
Identity matrix I
diagonal matrix with elements equal to 1
Symmetric matrix
square matrix where a[i][j]=a[j][i]
Sparse matrices
Many problems are based on matrices, which contains at most 1% of non-zero elements. It’s more efficient and faster if we store them in the SPARSE format.
e.g.: compressed row storage CRS
Arithmetic
Addition
Subtraction
Multiplication -> NON-commutative
Dot product
Transposition
At or A’
swapping rows and columns in such way that after it a[i][j] ==a[j][i]
(AB)t=BtAt
Fractal
Set with following features:
Benoit Mandelbrote
MANDELBROT fractal:
Norm of a vector
it’s length
Orthogonal vectors
Dot product = 0 <=> at * b = 0