[Depmix-commits] r408 - pkg/depmixS4

Wed Mar 10 17:07:00 CET 2010

Author: ingmarvisser
Date: 2010-03-10 17:07:00 +0100 (Wed, 10 Mar 2010)
New Revision: 408

Modified:
   pkg/depmixS4/DESCRIPTION
   pkg/depmixS4/NEWS
   pkg/depmixS4/README
Log:
Updated version nr (now at 1.0-0 !)

Modified: pkg/depmixS4/DESCRIPTION
===================================================================

--- pkg/depmixS4/DESCRIPTION	2010-03-10 16:05:13 UTC (rev 407)
+++ pkg/depmixS4/DESCRIPTION	2010-03-10 16:07:00 UTC (rev 408)
@@ -1,7 +1,7 @@
 Package: depmixS4
-Version: 0.9-0
-Date: 2010-03-08
-Title: Dependent Mixture Models
+Version: 1.0-0
+Date: 2010-03-10
+Title: Dependent Mixture Models in S4
 Author: Ingmar Visser <i.visser at uva.nl>, Maarten Speekenbrink <m.speekenbrink at ucl.ac.uk>
 Maintainer: Ingmar Visser <i.visser at uva.nl>
 Depends: R (>= 2.9.1), stats, nnet, methods, MASS, Rsolnp

Modified: pkg/depmixS4/NEWS
===================================================================
--- pkg/depmixS4/NEWS	2010-03-10 16:05:13 UTC (rev 407)
+++ pkg/depmixS4/NEWS	2010-03-10 16:07:00 UTC (rev 408)
@@ -1,14 +1,29 @@
 
-Changes in depmixS4 version 0.4-0
+Changes in depmixS4 version 1.0-0
 
   o added optimization using Rsolnp, which can be invoked by using 
-    method="rsolnp" in calling fit on (dep-)mix objects. The Rsolnp 
-    solver is currently available from r-forge. 
+    method="rsolnp" in calling fit on (dep-)mix objects. Note that 
+    this is meant for fitting models with additional constraints. 
+    Method="rsolnp" is now the default when fitting constrained 
+    models, but method="donlp" is still supported. 
 
   o added documentation for control arguments that can be passed to 
     em algorithm. 
 
+  o added multinomial models with identity link for transition and prior 
+    probabilities. These are now the default when no covariates are 
+    present. 
 
+  o added bounds and constraints for multinomial identity models such
+    that these constraints are satisfied when fitting models with 
+    method="rsolnp" or "donlp". Also, variance and sd parameters in 
+    gaussian and multivariate normal models are given bounds to 
+    prevent warnings and errors in optimization of such models using 
+    rsolnp or donlp. 
+
+  o added option to generate starting values as part of the EM 
+    algorithm. 
+
 Changes in depmixS4 version 0.3-0
   
   o added multinomial response function with identity link (no covariates

Modified: pkg/depmixS4/README
===================================================================
--- pkg/depmixS4/README	2010-03-10 16:05:13 UTC (rev 407)
+++ pkg/depmixS4/README	2010-03-10 16:07:00 UTC (rev 408)
@@ -1,17 +1,42 @@
 
-depmixS4 provides a framework for specifying and fitting hidden Markov models. Currently, it interfaces the glm functions to specify the state dependent measurement models. There is also a multinomial() family function that provides functionality for multinomial logistic responses with covariates. The transition matrix and the initial state probabilities are also modeled as multinomial logistics with the possibility of including covariates. Optimization is by default done by the EM algorithm. When linear constraints are included, Rdonlp2 is used for optimization (see details below). New response distributions can be added by extending the response-class and writing appropriate methods for it (dens, and getpars and setpars). depmixS4 also fits latent class and mixture models. 
+depmixS4 provides a framework for specifying and fitting hidden Markov
+models.  The observation densities use an interface to the glm
+distributions, most of which are now implemented.  Besides these,
+observations can be modelled using the multinomial distribution with
+identity or logistic link function.  The latter provides functionality for
+multinomial logistic responses with covariates.  The transition matrix and
+the initial state probabilities are also modeled as multinomial logistics
+(or multinomials with identity link) with the possibility of including
+covariates.  Optimization is by default done by the EM algorithm.  When
+linear constraints are included, package Rsolnp is used for optimization
+(there is also support for using Rdonlp2 as optimizer, see details below).
+New response distributions can be added by extending the response-class and
+writing appropriate methods for it (dens, and getpars and setpars); an
+example of this is provided on the ?makeDepmix help page.  depmixS4 also
+fits latent class and mixture models.
 
 The latest development version of depmix can be found at: 
 https://r-forge.r-project.org/projects/depmix/
 
 
-DIFFERENCES BETWEEN DEPMIXS4 AND DEPMIX
+FOR DEPMIX USERS
 
-depmixS4 is a completely new implementation of the depmix package using S4 classes. Model specification now uses formulae and family objects, familiar from the lm and glm functions. Moreover, the transition matrix and the initial state probabilities (as well as multinomial responses) are now modeled by default as multinomial logistics with a baseline. Specification of linear constraints uses the same mechanism as was used in depmix, with the only difference that constraints are passed as arguments to the fit function rather than the model specification function. See the help files for further details.
+depmixS4 is a completely new implementation of the depmix package using S4
+classes.  Model specification now uses formulae and family objects,
+familiar from the lm and glm functions.  Moreover, the transition matrix
+and the initial state probabilities (as well as multinomial responses) are
+now modeled as multinomial logistics with a baseline.  Specification of
+linear constraints uses the same mechanism as was used in depmix, with the
+only difference that constraints are passed as arguments to the fit
+function rather than the model specification function.  See the help files
+for further details.
 
 
-USING RDONLP2
-
-Optimization of models with (general) linear (in-)equality constraints can be done using the Rdonlp2 package, written Ryuichi Tamura(ry.tamura @ gmail.com), which is available from: http://arumat.net/Rdonlp2/
+USING RDONLP2
 
-Optimization with Rdonlp2 is automatically selected when constraints are specified in the fit function. 
+Optimization of models with (general) linear (in-)equality constraints can
+be done using the Rdonlp2 package, written Ryuichi Tamura(ry.tamura @
+gmail.com), which is available from: http://arumat.net/Rdonlp2/
+
+Optimization with Rdonlp2 is automatically selected when constraints are
+specified in the fit function.

