https://github.com/ElsevierSoftwareX/SOFTX_2019_219
Revision 41f675a3cc5f79705531204f93deb88aca428611 authored by Konrad Werys on 09 August 2019, 09:23:55 UTC, committed by Konrad Werys on 09 August 2019, 09:23:55 UTC
1 parent ba1145d
Tip revision: 41f675a3cc5f79705531204f93deb88aca428611 authored by Konrad Werys on 09 August 2019, 09:23:55 UTC
build: adding lmfit
build: adding lmfit
Tip revision: 41f675a
notepad.md
# Different results from the configurations:
parameter_to_map: T1_SHMOLLI
functions_type: **FunctionsBasic/FunctionsShmolli**
fitting_method: LevMarVnl
max_function_evals: 4000
fTolerance: 1e-12
sign_calc_method: MagPhase
start_point_calc_method: Default
FunctionsBasic: 10s and seemingly correct results
FunctionsShmolli: 2.2s and wrong results
parameter_to_map: T1_SHMOLLI
functions_type: FunctionsShmolli
fitting_method: LevMarVnl
max_function_evals: 4000
fTolerance: 1e-12
sign_calc_method: MagPhase
start_point_calc_method: **Default/StartPointSHMOLLI**
Default: 2.2s and wrong results
StartPointSHMOLLI: 10s and seemingly correct results
# 5 vs 7 samples
5 samples are calculated way slower. Maybe playing with fmax and maxIter would reduce the time?
Computing file changes ...