NMinimize vs NonlinearModelFit
- To: mathgroup at smc.vnet.net
- Subject: [mg114450] NMinimize vs NonlinearModelFit
- From: Alex <axel.kowald at rub.de>
- Date: Sat, 4 Dec 2010 06:17:08 -0500 (EST)
Hi everybody, I would like to fit some model parameter to numerical data and so NonlinearModelFit seems to be the right choice. It also gives nice confidence intervals and other statistics for the obtained parameter values, but then NMinimize can use the DifferentialEvolution algorithm, which is not available to NonlinearModelFit. I would also like to put constraints on my parameters (lower & upper limit), but it seems that both methods can do this. Basically I wonder what method other people are using for fitting more complex models (10-15 parameters) to numerical data ? axel