# nmsurv

Calculates max likelihood estimates using Nelder Mead's simplex method

## Contents

## Syntax

[q, info] = **nmsurv**(func, p, varargin)

## Description

Calculates max likelihood estimates using Nelder Mead's simplex method similar to nrsurv, but slower and a larger bassin of attraction

Input

- func: string with name of user-defined function

f = func (p, tn) with p: k-vector with parameters; tn: (n,c)-matrix; f: n-vector [f1, f2, ...] = func (p, tn1, tn2, ...) with p: k-vector and tni: (ni,k)-matrix; fi: ni-vector with model predictions The dependent variable in the output f; For tn see below.

- p: (k,2) matrix with

p(:,1) initial guesses for parameter values p(:,2) binaries with yes or no iteration (optional)

- tni (read as tn1, tn2, .. ): (ni,2) matrix with

tni(:,1) time: must be increasing with rows tni(:,2) number of survivors: must be non-increasing with rows tni(:,3, 4, ... ) data-pont specific information data (optional) The number of data matrices tn1, tn2, ... is optional but >0

Output

- q: matrix like p, but with ml-estimates
- info: 1 if convergence has been successful; 0 otherwise

## Remarks

Calls user-defined function 'func' Set options with html **nmregr_options** See **scsurv** for the definition of the user-defined function, and **scsurv2** and **nmsurv2** for 2 independent variables and **scsurv3** and **nmsurv3** for 3 independent variables. It is usually a good idea to run **scsurv** on the result of nmsurv.

## Example of use

See **mydata_surv**