Hello,
I've put the short version here and if anyone wants to run the code with
CollocInfer, I've given the full version in the file "analysis".
I come at the question of array attributes and dimnames
to try to simplify.
In a CollocInfer LS.profile analysis using this array 'Y' constructed
as follows:
w=rnorm(41,.05)
z=rnorm(41,.06)
yX<-cbind(w,z)
y<-as.array(yX)
colnames(y)=c("V","R")
Y<-array(0,c(41,2,2))
Y[,1,]=y
Y[,2,]=y
I receive an error
Error in as.array.default(Y) : attempt to set an attribute on NULL
So I think to name the attributes :
varnames=c("V","R")
rownames(Y)<-rownames(Y, do.NULL = FALSE, prefix = "Obs.")
colnames(Y)[1:2]=c("one","two")
assign("dimnames(Y)",list(rownames(Y),varnames,colnames(Y)))
Y2<-as.array(Y,dimnames=dimnames(Y))
assign("Y",Y2)
then rerun the analysis with the same result
Now traceback() gives:
7: as.array.default(Y)
6: as.array(Y)
5: kronecker(X, Y)
4: kronecker(X, Y)
3: diag(rep(1, nrep)) %x% basisvals$bvals.obs
2: LS.setup(pars, coefs, fn, basisvals, lambda, fd.obj, more, data,
weights, times, quadrature, eps = 1e-06, posproc, poslik,
discrete, names, sparse)
1: Profile.LS(fhn, data = data2, times = times, pars = pars, coefs = coefs,
lambda = lambda, out.meth = "nls", control.in = control.in,
control.out = control.out)
the first four numbers here(7..4) seem okay when I call each; but in number 3:
calling the given text produces the error:
error in evaluating the argument 'X' in selecting a method for
function 'kronecker': Error in diag(rep(1, nrep)) :
error in evaluating the argument 'x' in selecting a method for
function 'diag': Error: object 'nrep' not found
At the outset in the manual, Hooker refers to the diagonal matrix, it
seems without further explanation:
In order to demonstrate replicated observations, we make use of another
set of
data generated at di erent initial conditions. We then need
concatenate these ob-
servations in time, and create new values for bvals and weights. The
function
diag.block from the simex package is used below, but there are several
packages
in R that provide block-diagonal matrices.
I have a feeling this diagonal matrix is a component of R analysis
that, if corrected here,
could produce results, and I would be grateful if anyone who has
experience with its use
could offer some help.
A
Okay here I come at the question of attributes and dimnames
from another direction.
times = seq(0,20,0.5)
pars = c(0.2,0.2,3)
names(pars) = c('a','b','c')
x0 = c(-1,1)
names(x0) = c('V','R')
fhn = make.fhn()
w=rnorm(41,.05)
z=rnorm(41,.06)
yX<-cbind(w,z)
y<-as.array(yX)
colnames(y)=c("V","R")
#a perturbation:#
u=rnorm(41,.02)
v=rnorm(41,.02)
m13=w+u
m14=w+v
m15<-cbind(m13,m14)
data<-m15
knots = seq(0,20,0.2)
norder = 3
nbasis = length(knots) + norder - 2
range = c(0,20)
bbasis = create.bspline.basis(range=range, nbasis=nbasis, norder=norder,
breaks=knots)
fd.data = array(data,c(nrow(data),1,ncol(data)))
fd.data = array(data,c(nrow(data),1,ncol(data)))
varnames=c("V","R")
DEfd = Data2fd(fd.data, times, bbasis,
fdnames=list(NULL,NULL,varnames) )
coefs = DEfd$coefs
colnames(coefs) = varnames
control=list()
control$trace = 0
control$maxit = 1000
control$maxtry = 10
control$reltol = 1e-6
control$meth = "BFGS"
control.in = control
control.in$reltol = 1e-12
control.in$print.level = 0
control.in$iterlim = 1000
control.out = control
control.out$trace = 2
lambda = c(10000,10000)
res0 = Profile.LS(fhn,data,times,pars,coefs=coefs,basisvals=bbasis,
lambda=lambda,in.meth='nlminb',out.meth='ProfileGN',
control.in=control.in,control.out=control.out)
#(This is 'y' for an initial analysis of the FhN data.)#
#so far no problem: the arrays 'y' and 'data' are both acceptable
now we want to extend this to two replicates of the data to be analyzed:#
data2 = array(0,c(41,2,2))
m16<-rnorm(41,.2)
m20=cbind(m16,m16)
m21<-as.array(m20,dim=c(41,2))
m22<-rnorm(41,.1)
m20=cbind(m22,m22)
m23<-as.array(m20,dim=c(41,2))
data2[,1,]=m21+y
data2[,2,]=m23+y
fd.data2 = array(data2,c(nrow(data2),2,ncol(data2)))
DEfd2 = Data2fd(fd.data2, times, bbasis,
fdnames=list(NULL,NULL,varnames))
res3 = Profile.LS(fhn, data=data2, times=times, pars=pars,
coefs=coefs, lambda=lambda, out.meth='nls',
control.in=control.in, control.out=control.out)
#Rather than parameter estimates, as with the single replicate analysis, this
produces the error:
Error in as.array.default(Y) : attempt to set an attribute on NULL
So thinking that the dimnames of either array are the "NULL" that R is
referring to,
I supply dimnames to the two arrays involved, data2, and y:#
rownames(data2)<-rownames(data2, do.NULL = FALSE, prefix = "Obs.")
colnames(data2)[1:2]=c("one","two")
dimnames(data2)=list(rownames(data2),colnames(data2),varnames)
data2a<-as.array(data2,dim=c(41,2,2),dimnames=dimnames(data2))
assign("data2",data2a)
Y<-array(0,c(41,2,2))
Y[,1,]=y
Y[,2,]=y
#Then rerun the analysis with same result.
Now traceback() gives:#
7: as.array.default(Y)
6: as.array(Y)
5: kronecker(X, Y)
4: kronecker(X, Y)
3: diag(rep(1, nrep)) %x% basisvals$bvals.obs
2: LS.setup(pars, coefs, fn, basisvals, lambda, fd.obj, more, data,
weights, times, quadrature, eps = 1e-06, posproc, poslik,
discrete, names, sparse)
1: Profile.LS(fhn, data = data2, times = times, pars = pars, coefs =
coefs,
lambda = lambda, out.meth = "nls", control.in = control.in,
control.out = control.out)
#the first four numbers here(7..4) seem okay when I call each; but in number 3:
calling the given text produces the error:
error in evaluating the argument 'X' in selecting a method for function
'kronecker': Error in diag(rep(1, nrep)) :
error in evaluating the argument 'x' in selecting a method for function
'diag': Error: object 'nrep' not found
At the outset in the manual, Hooker refers to the diagonal matrix, it seems
without further explanation:
In order to demonstrate replicated observations, we make use of another
set of
data generated at di erent initial conditions. We then need concatenate
these ob-
servations in time, and create new values for bvals and weights. The
function
diag.block from the simex package is used below, but there are several
packages
in R that provide block-diagonal matrices.
I have a feeling this diagonal matrix is a component of R analysis that, if
corrected here,
could produce results, and I would be grateful if anyone who has experience
with its use
could offer some help.
A
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
