A
A
Add **robust regression** implicit sublinear function to callbacks. It seems fairly easy to attach to the already existing callbacks a hook to add support for implicitly adding robustness functions. Added...
# Add support for derivative checker The idea is to add and option (`options%check_derivatives = .True.`) that checks for: 1. 1st order derivatives (user Jacobian vs FD approximation) 2. 2nd...
Add support to a new debug build using NAGFOR and all debug features, mainly to build the RALFit package with the flags: ```{bash} nagfor -g -nan -C=all -C=undefined -u -ieee=full...
The docs has one error in the description for the user Call-back `eval_J` # Ordering of the Jacobian elements Jacobian of the residuals is a matrix $J$ of $m$ rows...
# Tensor Newton subproblem solve (with regularization) ## Part 1 of 2 Nan's under valid options in recursive call (regularization) In the recursive call (model=4) the starting point is either...
# Subroutine `AINT_TR` 1. Deal separately with B generic and B=Identinity, I(n). 2. w%B is acting as a dense Identity matrix, and is mostly not required, except for the calls...
# Under the context of Tensor-Newton Subproblem. ## Multiple redundant calls to `eval_HF` when `eval_HP` is present ### Problem In the recursive call to `evaltensor_hf` there is a redundant call...
Example program `Lanczos.f90` causes exception `Warning: Floating divide by zero occurred` in call to ``` call dgesvd( JOBU, JOBVT, n, m, w%Jcopy, n, w%S, w%S, 1, w%S, 1, & ```...
Need to make sure `regularization_solve` is NOT called with `opt%reg_order /= 2`. We should catch the specific combination of options that triggers this "not yet supported" combination before entering the...
Consider removing the following unused parameters: ``` src/ral_nlls_workspaces.f90, line 998: Unused dummy variable OPTIONS src/ral_nlls_workspaces.f90, line 1109: Unused dummy variable OPTIONS src/ral_nlls_workspaces.f90, line 1316: Unused dummy variable OPTIONS src/ral_nlls_workspaces.f90, line...