matrix_svd_exists
ProvedExistence of a (reduced) singular value decomposition for an arbitrary real matrix. For every real matrix there exist a rank and an SVD datum — that is, strictly positive singular values together with orthonormal left vectors and orthonormal right vectors such that . This is the classical reduced SVD existence theorem: is the number of nonzero singular values, the are an orthonormal eigenbasis of the Gram operator restricted to its positive eigenspaces (with the corresponding eigenvalue), and . The platform SVD structure carries exactly this data (with sigma_pos, u_orthonormal, v_orthonormal, decomp). Proof: diagonalize the symmetric positive operator by the spectral theorem to obtain orthonormal right singular vectors and eigenvalues ; set , keep only the indices with , define (orthonormal since ), and verify from the eigenbasis reconstruction (terms with have ).
import Definitions.Def_matrix_completion_svd open MatrixCompletion
theorem matrix_svd_exists {n1 n2 : Nat} (N : Matrix (Fin n1) (Fin n2) ℝ) : ∃ (r : Nat) (S : SVD N r), True := by sorry