# [Pomp-commits] r556 - in pkg: . inst man

Fri Sep 2 16:02:32 CEST 2011

Author: kingaa
Date: 2011-09-02 16:02:32 +0200 (Fri, 02 Sep 2011)
New Revision: 556

Modified:
pkg/DESCRIPTION
pkg/inst/NEWS
pkg/man/mif.Rd
pkg/man/pfilter.Rd
Log:

- improve the help files for 'pfilter' and 'mif' so as to explain how to use non-scalar (vector or function) values for the 'Np' argument.
- the 'mif' help file description of the 'ic.lag' argument also improved

Modified: pkg/DESCRIPTION
===================================================================
--- pkg/DESCRIPTION	2011-08-31 14:30:25 UTC (rev 555)
+++ pkg/DESCRIPTION	2011-09-02 14:02:32 UTC (rev 556)
@@ -1,8 +1,8 @@
Package: pomp
Type: Package
Title: Statistical inference for partially observed Markov processes
-Version: 0.39-2
-Date: 2011-08-31
+Version: 0.39-3
+Date: 2011-09-02
Author: Aaron A. King, Edward L. Ionides, Carles Breto, Steve Ellner, Bruce Kendall, Helen Wearing, Matthew J. Ferrari, Michael Lavine, Daniel C. Reuman
Maintainer: Aaron A. King <kingaa at umich.edu>
URL: http://pomp.r-forge.r-project.org

Modified: pkg/inst/NEWS
===================================================================
--- pkg/inst/NEWS	2011-08-31 14:30:25 UTC (rev 555)
+++ pkg/inst/NEWS	2011-09-02 14:02:32 UTC (rev 556)
@@ -1,4 +1,7 @@
NEWS
+0.39-3
+     o	The help files for 'pfilter' and 'mif' have been updated to explain how to use a variable number of particles in the particle filtering algorithm.
+
0.39-2
o	When 'po' is a 'pomp'-class object with covariates, 'as.data.frame(po)' and 'as(po,"data.frame")' now contain columns for the covariates.
The latter are interpolated, if necessary, at the observation times.

Modified: pkg/man/mif.Rd
===================================================================
--- pkg/man/mif.Rd	2011-08-31 14:30:25 UTC (rev 555)
+++ pkg/man/mif.Rd	2011-09-02 14:02:32 UTC (rev 556)
@@ -72,12 +72,19 @@
the name of every positive element of \code{rw.sd} must be in either \code{pars} or \code{ivps}.
}
\item{Np}{
-    a positive integer;
-    the number of particles to use in filtering
+    the number of particles to use in filtering.
+    This may be specified as a single positive integer, in which case the same number of particles will be used at each timestep.
+    Alternatively, if one wishes the number of particles to vary across timesteps, one may specify \code{Np} either as a vector of positive integers (of length \code{length(time(object,t0=TRUE))}) or as a function taking a positive integer argument.
+    In the latter case, \code{Np(k)} must be a single positive integer, representing the number of particles to be used at the \code{k}-th timestep:
+    \code{Np(0)} is the number of particles to use going from \code{timezero(object)} to \code{time(object)[1]},
+    \code{Np(1)}, from \code{timezero(object)} to \code{time(object)[1]},
+    and so on, while when \code{T=length(time(object,t0=TRUE))},
+    \code{Np(T)} is the number of particles to sample at the end of the time-series.
}
\item{ic.lag}{
a positive integer;
-    the timepoint for fixed-lag smoothing of initial-value parameters
+    the timepoint for fixed-lag smoothing of initial-value parameters.
+    The \code{mif} update for initial-value parameters consists of replacing them by their filtering mean at time \code{times[ic.lag]}, where \code{times=time(object)}.
}
\item{var.factor}{
a positive number;

Modified: pkg/man/pfilter.Rd
===================================================================
--- pkg/man/pfilter.Rd	2011-08-31 14:30:25 UTC (rev 555)
+++ pkg/man/pfilter.Rd	2011-09-02 14:02:32 UTC (rev 556)
@@ -32,7 +32,14 @@
If it desired that all particles should share the same parameter values, one one may supply \code{params} as a named numeric vector.
}
\item{Np}{
-    integer; number of particles to use.
+    the number of particles to use.
+    This may be specified as a single positive integer, in which case the same number of particles will be used at each timestep.
+    Alternatively, if one wishes the number of particles to vary across timesteps, one may specify \code{Np} either as a vector of positive integers (of length \code{length(time(object,t0=TRUE))}) or as a function taking a positive integer argument.
+    In the latter case, \code{Np(k)} must be a single positive integer, representing the number of particles to be used at the \code{k}-th timestep:
+    \code{Np(0)} is the number of particles to use going from \code{timezero(object)} to \code{time(object)[1]},
+    \code{Np(1)}, from \code{timezero(object)} to \code{time(object)[1]},
+    and so on, while when \code{T=length(time(object,t0=TRUE))},
+    \code{Np(T)} is the number of particles to sample at the end of the time-series.
When \code{object} is of class \code{mif}, this is by default the same number of particles used in the \code{mif} iterations.
}
\item{tol}{