Opened 14 years ago
Closed 13 years ago
#160 closed feature (fixed)
include genetic optimization as an option for curve fitting
| Reported by: | srkline | Owned by: | srkline |
|---|---|---|---|
| Priority: | major | Milestone: | Analysis Wish List |
| Component: | Analysis | Keywords: | |
| Cc: | Blocking: | ||
| Task: |
Description
Include Andy Nelson's XOP as an option for genetic optimization. Safest to distribute it with our package, could easily check for its existence before calling it...
Change History (3)
comment:1 Changed 14 years ago by srkline
comment:2 Changed 14 years ago by srkline
Works in principle after Andy's fix, but needs to be fully incorporated. Straightforward, but tedious. Also, Genetic Opt is VERY SLOW. OK for simple models, but almost intolerably slow for even simple smeared models.
comment:3 Changed 13 years ago by srkline
- Resolution set to fixed
- Status changed from new to closed
Added a procedure file that does the necessary wrapping for GenCurveFit?, since it's not a perfect drop-in replacement for FuncFit?. r495
Switch between regular L-M and GenOp? using a menu item.
Still ridiculously slow to use.
Can't yet be used with global fitting. Would be a real pain to implement. Can't imagine how slow that would be to use...
Looks like we can implement this with a wrapper, but as of 10/2, there was an error in Andy's XOP for structure fits. e-mailed him for a solution.