You want to turn arrays of arrays of numbers into mathematical matrices, and multiply the matrices together.
You can create
Matrix objects from arrays of arrays, and
multiply them together with the
require ' matrix' require 'mathn' a1 = [[1, 1, 0, 1], [2, 0, 1, 2], [3, 1, 1, 2]] m1 = Matrix[*a1] # => Matrix[[1, 1, 0, 1], [2, 0, 1, 2], [3, 1, 1, 2]] a2 = [[1, 0], [3, 1], [1, 0], [2, 2.5]] m2 = Matrix[*a2] # => Matrix[[1, 0], [3, 1], [1, 0], [2, 2.5]] m1 * m2 # => Matrix[[6, 3.5], [7, 5.0], [11, 6.0]]
Note the unusual syntax for creating a
Matrix object: you pass the rows of the
matrix into the array indexing operator, not into
Matrix#new (which is private).
Matrix class overloads
the arithmetic operators to support all the basic matrix arithmetic
operations, including multiplication, between matrices of compatible
dimension. If you perform an arithmetic operation on incompatible
matrices, you'll get an
Multiplying one matrix by another is simple enough, but multiplying a chain of matrices together can be faster or slower depending on the order in which you do the multiplications. This follows from the fact that multiplying a matrix with dimensions K x M, by a matrix with dimensions MxN, requires K * M * N operations and gives a matrix with dimension K * N. If K is large for some matrix, you can save time by waiting til the end before doing multiplications involving that ...