nlsrSS function

nlsrSS - solve selfStart nonlinear least squares with nlsr package

This function uses the getInitial() function to estimate starting parameters for a Gauss-Newton iteration, then calls nlsr::nlxb() appropriately to find a solution to the required nonlinear least squares problem.

nlsrSS(formula, data)

Arguments

• formula: a model formula incorporating a selfStart function in the right hand side
• data: a data frame with named columns that allow evaluation of the formula

Returns

A solution object of class nlsr.

List of solution elements

• resid: weighted residuals at the proposed solution

• jacobian: Jacobian matrix at the proposed solution

• feval: residual function evaluations used to reach solution from starting parameters

• jeval: Jacobian function (or approximation) evaluations used

• coefficients: a named vector of proposed solution parameters

• ssquares: weighted sum of squared residuals (often the deviance)

• lower: lower bounds on parameters

• upper: upper bounds on parameters

• maskidx: vector if indices of fixed (masked) parameters

• weights: specified weights on observations

• formula: the modeling formula

• resfn: the residual function (unweighted) based on the formula

Author(s)

J C Nash 2022-9-14 nashjc at uottawa.ca