Take A = (1 1; 0 0). The first link multiplies it by itself and the A × B row returns (1 1; 0 0): the same matrix, so A² = A, which is what 'projection' means algebraically — do it once and doing it again changes nothing. The second link is A alone with its two diagnostics, determinant 0 and trace 1. The third link is the entry that has to come back, 1·1 + 1·0 = 1, and the fourth is the cleanest projection of the lot, onto the x-axis, (1 0; 0 0).
The determinant has to be 0: A² = A gives A(A − I) = 0, so an idempotent other than the identity has a kernel, and here the kernel is the line spanned by (1, −1). The trace is then forced to equal the rank, because the only possible eigenvalues of an idempotent are 0 and 1, each occurring with its own multiplicity, and the trace counts the ones. So the pair (det 0, trace 1) says 'rank-one projection' as compactly as anything can, and it says it at every size: a 4×4 idempotent of rank 3 has trace 3. This is the calculation that makes the trace more than a sum of diagonal entries, and the reason 'the trace of a projection is the dimension of the subspace' survives generalisation to operators on infinite-dimensional spaces and to the K-theory of idempotents.
Answer
0
How to read it
The determinant of the idempotent. The first link squares the matrix, the second prints determinant and trace, the third one entry of the square, the fourth the projection onto the x-axis.
Source
Idempotent matrices are the algebraic definition of a projection; the orthogonal projection onto a line through the origin, P = uuᵀ/(uᵀu), is the standard construction, worked in this form in Gilbert Strang's 'Linear Algebra and Its Applications' (1976)
Work it in the calculator
These links open the right calculator with this problem's numbers already typed in. The line above is what the screen should show — check it yourself.