Current voxels branch#559

Open

akaptano wants to merge 115 commits intomasterfrom

Commits on Jan 10, 2023

Made trivial change so can save this as a new simsopt branch.
akaptano
committed

Commits on Jan 11, 2023

Got the flux constraint matrix up and running to an extent. Trying to do some tests to see if it is working, like checking that projection onto C * alpha actually produces a divergenceless current.…
akaptano
committed
Took a stab at reducing and fixing the grid generation. The grid generation actually looked pretty good a day or two ago and now it looks bad so Im not sure what went wrong. I think it was only wor…
akaptano
committed

Commits on Jan 12, 2023

Commits on Jan 13, 2023

Commits on Jan 17, 2023

Gotten very close to getting the bnormal part of the optimization working. the least squares calculation of Bnormal actually appears to be correct, as far as I can tell, but for some reason the Bno…
akaptano
committed

Commits on Jan 18, 2023

Commits on Jan 19, 2023

Commits on Jan 20, 2023

Commits on Jan 21, 2023

Commits on Jan 23, 2023

Think I mostly have the baseline current voxel method working now, with all the L2 loss terms and the flux jumps working. A few things to do still: (1) investigate mysterious blow up in the optimiz…

user01
authored and
user01
committed

Commits on Jan 25, 2023

Commits on Jan 26, 2023

Made a number of bug fixes in the grid connection calculations (like dxx <= dx does not work if they are equal, this is classic floating point error), and hopefully have that finally working right …

user01
authored and
user01
committed

Commits on Jan 27, 2023

Fixed some grid issues. Added XSIMD of the B and Bext calculations, which seems to mildly speed things up (factor of 2 ish). Tried coding up accelerated prox gradient descent in c++ but not seeing …
akaptano
committed

Commits on Jan 31, 2023

Tried a bunch of different things for speedup. Started by removing the svd, which inevitably gets long as matrices get larger. Instead, there is nice upper bound estimate for the largest singular v…

user01
authored and
user01
committed

Commits on Feb 1, 2023

Commits on Feb 3, 2023

Trying to get the WindingVolume class to save properly so that I can call the poincare plots on it in postprocessing script but having issues so going to merge in main branch before continuing on.

user01
authored and
user01
committed
Merge branch 'master' into winding_volume

user01
authored and
user01
committed
Added the relax and split piece but not sure yet if the l0 part is really working as desired. Still havent been able to figure out the poincare plots.

user01
authored and
user01
committed

Commits on Feb 10, 2023

made some modest changes to the optimization. gearing up for parameter scans and other checks before getting too deep into trying to get the current voxel solutions to look just right.

user01
authored and
user01
committed
commiting current script because not seeing openmp speedups on my laptop, and want to check on nersc if this is just an issue on my laptop.

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Feb 15, 2023

Tried to get Itarget working as one of the constraints, but causes issues with the accelerated prox grad algorithm I am using because it is of the form Cx = d instead of Cx = 0. Going to try a litt…

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Feb 16, 2023

Trying to code up the exact KKT solution for the method. However, this and all previous methods have been stumped somewhat by the constraint matrix not being full rank, so this is going to be the f…

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Feb 17, 2023

Trying, so far unsuccessfully, to figure out why the constraint matrix is not full rank. The rows (corresponding to individual constraints) appear to be linearly inedpendent, but the columns (corre…

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Feb 21, 2023

Working on getting the analytic and projected gradient descent methods both correct and comparable. Pseudoinverse for the constraint matrix inversion for the projection operator appears to work oka…

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Feb 22, 2023

After trying to get the hard Itarget constraint working for a while, concluded that this is too painful to justify not doing the easy thing and using it as a loss term. Also KKT version is quite sl…

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Mar 1, 2023

Working on getting a realistic example working for an axisymmetric tokamak. Saving current progress to try running it faster on NERSC.

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed

Commits on Mar 2, 2023

Got up to Nx = 50 runs going pretty well on haswell. Fixed semimajor bug, which is that the scaling for Itarget matrix was off (sqrt(nphi) instead of the correct nphi) so optimizations were using a…
akaptano
committed

Commits on Mar 3, 2023

Played around a bit more with the axisymmetric example. matt and I discussed how Itarget loss term doesnt seem to be matched visually but appears to be calculated correctly in the code. One possibl…
akaptano
committed

Commits on Mar 4, 2023

Added routine to save voxels in vtk format
landreman
committed

Commits on Mar 6, 2023

Commits on Mar 7, 2023

Added in Matts plotting functionality, shifted the grid by half a grid spacing (so that there is no gap when it gets repeated around the torus), shifted the polynomial basis functions by their midp…
akaptano
committed

Commits on Mar 8, 2023

Ran the QA example without sparsity. Saving these changes so I can switch branches.
akaptano
committed

Commits on Apr 6, 2023

Added the beginning of a QH run.

Alan Kaptanoglu
authored and
Alan Kaptanoglu
committed
Merge branch 'winding_volume' of https://github.com/hiddenSymmetries/simsopt into winding_volume
akaptano
committed

Commits on Apr 15, 2023

added some functionality and got first version of constrained stlsq working, which could be useful here.
akaptano
committed

Commits on Apr 17, 2023

Tried the cstlsq stuff out a bit more but doesnt seem very promising. Issue is that grid cells can chopped forever, and there is no recovery. Maybe this is fine on larger grids, but currently limit…
akaptano
committed

Commits on Apr 18, 2023

Commits on Apr 20, 2023

Working on getting the axisymmetric results really polished. Seem to be making some headway taking lots of very small threshold steps.
akaptano
committed

Commits on Apr 21, 2023

Commits on Apr 25, 2023

Played around with the tokamak and QA scripts until I got some decent looking solutions. Next step is to do the same with QH.
akaptano
committed

Commits on Apr 26, 2023

Merged with permanent magnet branch recent changes.
akaptano
committed

Commits on May 11, 2023

Committing current version of the repo. Unfortunately it is currently a mess, because of trying to debug the issue Matt and I identified this week, which seems to be spurious matching of the Itarge…
akaptano
committed

Commits on May 17, 2023

Commits on May 20, 2023

Commits on May 23, 2023

Getting MINRES incorporated into the full optimization loop. Initial results look fairly promising -- unclear yet if they outperform the accel. proj. grad descent methods yet but my guess is that t…
akaptano
committed

Commits on May 24, 2023

Commits on Jun 6, 2023

Saving current state of the code that is able to generate the paper figures. Lots of future work to clean up the code and document everything. Tried to code up first attempt at initializing coil ob…
akaptano
committed

Commits on Jun 9, 2023

Commits on Jun 12, 2023

Committing for now because perlmutter has been offline for days. going to go work on API documentation for the recent merge of permanent magnets stuff.
akaptano
committed

Commits on Jun 13, 2023

Perlmutter somehow down again. Switching to finishing up docs.
akaptano
committed

Commits on Jun 14, 2023

Cant find a bug so reverting and then coming back to fix this.
akaptano
committed

Commits on Jun 20, 2023

Commits on Jun 21, 2023

Commits on Sep 18, 2023

Committing current updates to the voxels branch, which were made a while ago.
akaptano
committed

Commits on Sep 22, 2023

Merged with current version of main branch since its been a while.
akaptano
committed

Commits on Oct 18, 2023

Commits on Oct 21, 2025

Commits on Oct 22, 2025

Commits on Oct 23, 2025

Commits on Oct 24, 2025

Commits on Feb 2, 2026

Commits on Feb 3, 2026

Commits on Feb 19, 2026

Commits on Feb 21, 2026

Fix coverage and remove tiny changes with main.
akaptano
committed

Commits on Mar 1, 2026

Merge branch 'master' into current_voxels
akaptano
committed

Commits on Mar 23, 2026