Add option to run finite difference on gradient of energy #52
Conversation
Codecov ReportAttention:
Additional details and impacted files@@ Coverage Diff @@
## main #52 +/- ##
==========================================
+ Coverage 88.52% 88.62% +0.10%
==========================================
Files 46 46
Lines 1839 1882 +43
==========================================
+ Hits 1628 1668 +40
- Misses 211 214 +3
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
non-linear-solver-spec.json
Outdated
| "doc": "Expensive Option: Eps for finite difference to verify gradient of energy." | ||
| }, | ||
| { | ||
| "pointer": "/advanced/gradient_fd_components", |
non-linear-solver-spec.json
Outdated
| @@ -319,5 +322,23 @@ | |||
| "default": 100, | |||
| "type": "int", | |||
| "doc": "Dangerous Option: Quit the optimization if the solver reduces the energy by less than f_delta for consecutive f_delta_step_tol steps." | |||
| }, | |||
| { | |||
| "pointer": "/advanced/gradient_fd", | |||
There was a problem hiding this comment.
maybe this instead of a bool is a string with 3 opts? none, full and directional?
src/polysolve/nonlinear/Solver.hpp
Outdated
| // Finite Difference Utilities | ||
| // ==================================================================== | ||
|
|
||
| bool gradient_fd = false; |
There was a problem hiding this comment.
all of these should be private and filed should start with m_
|
this pr contains the changes of the other pr too |
No description provided.