Open FHell opened 1 year ago
Incidentally: lsmr! and lsmr show the same behavior.
Edit:
And a nontrivial example:
x = ones(4)
b = ones(4)
v1 = rand(4)
v2 = rand(4)
v3 = rand(4)
w1 = rand(4)
w2 = rand(4)
w3 = rand(4)
A = v1 * transpose(w1) + v2 * transpose(w2) + v3 * transpose(w3)
#-
lsqr!(x, A, b)
y = lsqr(A, b)
#-
isapprox(x, y) # false
According to the documentation lsqr! should return the minimum norm solution if the matrix is ambiguous. This doesn't seem to work:
lsqr is working as expected:
Julia 1.8, IterativeSolvers v0.9.2