- nls2 Examples


The examples described in the book by S. Huet, and all. are provided in the package. References to the paragraphes of this book are given below.
Some additional examples are available to illustrate some particular uses of nls2.
The model files and the data-creation commands files are in the directory data of the package. The execution-commands are in the directory demo. Names of the files are: <example-name>.R

See also the examples list sorted by function-names.

The examples are:

Name in chapters illustrated functions characteristics
beetles 6.7: Binomial non-linear models nls2 introduced in the 2nd edition
numerical equalities constraints
cheese 7.1.8: Multinomial non-linear models nls2, confidence. introduced in the 2nd edition
corti 1.6: Nonlinear regression model and parameter estimation nls2 numerical equalities constraints
2.6: Accuracy of estimators, confidence intervals and tests confidence
4.6: Diagnostics of model misspecification nls2, plres, plvar
5.6: Calibration and Prediction calib
7.2.4: Poisson non-linear models nls2
ej   nls2 This example works when the Nag library is available, only. The loadnls2 command should be customised according to the localisation of the Nag library on the site.
The programs are not automatically generated but written "at hand" and use Nag subroutines. This is an example of estimation by three models with common parameters. Each model is calculated on a particular subset of data: two are calculated by C-programs which can be automatically generated, the third one is calculated by a Fortran program calling Nag subroutines.
Values are transmitted to these programs via Fortran COMMON.
elisa 1.6: Nonlinear regression model and parameter estimation pldnls2, nls2 with curves,
equality constraints between parameters and 'binf' or 'bsup' bounds
2.6: Accuracy of estimators, confidence intervals and tests wald, nls2, confidence, conflike, plfit, bootstrap
fbeetles 6.7: Binomial non-linear models nls2, confidence. introduced in the 2nd edition
with curves, numerical equality constraints between parameters
insect 6.7: Binomial non-linear models nls2, confidence. introduced in the 2nd edition
'binf' or 'bsup' bounds on parameters
isomer 1.6: Nonlinear regression model and parameter estimation nls2, plfit 'binf' or 'bsup' bounds on parameters
2.6: Accuracy of estimators, confidence intervals and tests confidence, ellips, iso, nls2, bootstrap
4.6: Diagnostics of model misspecification plfit, plres.
mag   nls2, renls2. bootstrap estimation on the Cortisol example. Works when the Nag library is available, only.
The model is written in C and calls Fortran subroutines from the Nag library to generate pseudo-observations.
miners 7.1.8: Multinomial non-linear models nls2, confidence. introduced in the 2nd edition
nasturnium 5.6: Calibration and Prediction nls2, bootstrap, calib  
orobranche 6.7: Binomial non-linear models nls2, bootstrap. introduced in the 2nd edition
ovo 1.6: Nonlinear regression model and parameter estimation nls2, plfitwith curves, model described by a odes, numerical equality constraints between parameters
2.6: Accuracy of estimators, confidence intervals and tests ellips, iso.
4.6: Diagnostics of model misspecification plres, plvar.
ownmethod   nls2. numerical equality constraints between parameters and 'binf' or 'bsup' bounds.
Example of an estimation method defined by the user.
You can use this method in case of errors auto-correlation. This file contains two examples: example of simulated data and example of real data. (Reference).
pasture 1.6: Nonlinear regression model and parameter estimation pldnls2, nls2, plfit  
2.6: Accuracy of estimators, confidence intervals and tests confidence, nls2, bootstrap
4.6: Diagnostics of model misspecification nls2, plfit.
5.6: Calibration and Prediction confidence, bootstrap.
pept 3.6: Variance estimation nls2, pldnls2, plres, plfit, conflike, confidence. alternated method of estimation
4.6: Diagnostics of model misspecification nls2, plres, plfit.  
perm   nls2, calcmodnls2. with curves and weights,
model described by a odes
pg   nls2, pldnls2, plres, plit, plfit. with curves, numerical equality constraints between parameters
root 4.6: Diagnostics of model misspecification pldnls2, nls2, plres, plfit.  
sim 4.6: Diagnostics of model misspecification nls2, confidence, plfit.  
simul   nls2, calcmodnls2. model calculated by a Fortran program. The regression function is solution of an implicit equation. Works when the Nag library is available, only.
substrat   nls2, calcmodnls2. with curves,
model described by a odes
numerical and equality constraints between parameters and 'binf' or 'bsup' bounds
tiller 3.6: Variance estimation nls2, confidence, conflike, bootstrap. changed in the 2nd edition
vaso 6.7: Binomial non-linear models nls2. introduced in the 2nd edition
equality constraints between parameters
explicit 'family': Bernoulli
volterra   nls2, calcmodnls2. model described by a odes
weibull   nls2, rexnls2.  

Reference for the example ownmethod:
D. Deschamps,
Rapport de Stage DESS Mathématiques Appliquées,
INRA, Jouy-en-Josas, juin 1994.

Sommaire


Creation: 2003 -