--fit-method FIT_METHOD
Sherpa fit method (simplex|moncar|levmar)
--fit-method
The default fit method is simplex which is a good compromise between speed and completeness. For the fastest fitting use levmar. If already have somewhat decent parameters and want to try to refine for the very best fit then select moncar. However, do not choose this option with more than about 10 or 15 free parameters as it can take a long time. Typically with moncar you need to start the fitting and then do something else for a while (many hours or more).
Neldermead ought to be referenced in the doc even if in SHERPA, neldermead
and simplex are aliases of the same thing.
While running a fit via gui_fit.py, you will be told that the "Method" is neldermead:
Dataset = 1 Method = neldermead Statistic = userstat Initial fit statistic = 220842 Final fit statistic = 66676.6 at function evaluation 1000 Data points = 1 Degrees of freedom = -22 Change in statistic = 154165 xijamod.heatsink3TSMXSPTT -47.298
xijamod.heatsink3TSMXSPTtau 7.10448
xijamod.prop_heat3TSMXSPTk 0.110314
..................
But neldermead is not referenced in your documentation:
http://cxc.cfa.harvard.edu/mta/ASPECT/tool_doc/xija/tutorial.html#fitting-a-model
--fit-method FIT_METHOD Sherpa fit method (simplex|moncar|levmar)
--fit-method The default fit method is simplex which is a good compromise between speed and completeness. For the fastest fitting use levmar. If already have somewhat decent parameters and want to try to refine for the very best fit then select moncar. However, do not choose this option with more than about 10 or 15 free parameters as it can take a long time. Typically with moncar you need to start the fitting and then do something else for a while (many hours or more).
Neldermead ought to be referenced in the doc even if in SHERPA, neldermead and simplex are aliases of the same thing.