GSL Multi-Start Nonlinear Least-Squares Fitting
Calculate variance-covariance matrix
Anova tables
Extract model coefficients
Confidence interval for model parameters
Confidence intervals for derived parameters
Model summary
Confidence intervals for derived parameters
Model deviance
Residual degrees-of-freedom
Extract model fitted values
Extract model formula
GSL Nonlinear Least Squares fitting
Tunable Nonlinear Least Squares iteration parameters
GSL Large-scale Nonlinear Least Squares fitting
Calculate leverage values
Extract model log-likelihood
Available NLS test problems
Retrieve an NLS test problem
Extract the number of observations
Calculate model predicted values
Extract model residuals
Residual standard deviation
An R interface to weighted nonlinear least-squares optimization with the GNU Scientific Library (GSL), see M. Galassi et al. (2009, ISBN:0954612078). The available trust region methods include the Levenberg-Marquardt algorithm with and without geodesic acceleration, the Steihaug-Toint conjugate gradient algorithm for large systems and several variants of Powell's dogleg algorithm. Multi-start optimization based on quasi-random samples is implemented using a modified version of the algorithm in Hickernell and Yuan (1997, OR Transactions). Robust nonlinear regression can be performed using various robust loss functions, in which case the optimization problem is solved by iterative reweighted least squares (IRLS). Bindings are provided to tune a number of parameters affecting the low-level aspects of the trust region algorithms. The interface mimics R's nls() function and returns model objects inheriting from the same class.