core.pseudoinverse

core.pseudoinverse.pseudoinverse(A)[source]

Calculate the pseudoinverse of array A over the last 2 axes (broadcasting the first axes) A* = ((A’.A)^(-1)).A’ where X’ is the transpose of X and X^-1 is the inverse of X

shapes: A: […,i,j]: A*: […,j,i]

core.pseudoinverse.weighted_pseudoinverse(A, W)[source]: Calculate the weighted pseudoinverse of array A over the last 2 axes (broadcasting the first axes) W is the weight matrix (diagonal) A* = ((A’.W.A)^(-1)).A’.W