We can multiply vectors by scalars, add, and subtract. Given vectors u, v and w we can form the linear combination x1u+x2v+x3w=b.
An example in R3 would be:
u=1−10,v=01−1,u=001
The collection of all multiples of u forms a line through the origin. The collection of all multiples of v forms another line. The collection of all combinations of u and v forms a plane. Taking all combinations of some vectors creates a subspace.
We could continue like this, or we can use a matrix to add in all multiples of w.
equals the sum x1u+x2v+x3w=b. The product of a matrix and a vector is a combination of the columns of the matrix. (This particular matrix A is a difference matrix because the components of Ax are differences of the components of that vector.)
When we say x1u+x2v+x3w=b we’re thinking about multiplying numbers by vectors; when we say Ax=b we’re thinking about multiplying a matrix (whose columns are u, v and w) by the numbers. The calculations are the same, but our perspective has changed.
For any input vector x, the output of the operation “multiplication by A” is some vector b:
A149=135
A deeper question is to start with a vector b and ask “for what vectors x does Ax=b?” In our example, this means solving three equations in three unknowns. Solving:
We see that x1=b1 and so x2 must equal b1+b2. In vector form, the solution is:
x1x2x3=b1b1+b2b1+b2+b3
But this just says:
x=111011001b1b2b3
or x=A−1b. If the matrix A is invertible, we can multiply on both sides by A−1 to find the unique solution x to Ax=b. We might say that A represents a transform x→b that has an inverse transform b→x.
In particular, if b=000 then x=000
The second example has the same columns u and v and replaces column vector w:
and our system of three equations in three unknowns becomes circular.
Where before Ax=0 implied x=0, there are non-zero vectors x for which Cx=0. For any vector x with x1=x2=x3,Cx=0. This is a significant difference; we can’t multiply both sides of Cx=0 by an inverse to find a nonzero solution x.