Nothing
R/coefKinresp.R
In twKinresp: Fitting kinetic models to microbial respiration data
Defines functions
coefKinrespMatrix
confintKinresp
.coefKinrespLogStart
coefKinrespNormStart
Documented in
coefKinrespMatrix
coefKinrespNormStart
confintKinresp
setMethodS3("kinrespParDist","gnls", function(
### Expected value and confidence interval of microbial parameters
model ##<< the result of the \code{\link{fitKinrespReplicate}} or \code{\link{fitKinrespExperiment}$model}
,... ##<< currently not used
){
# kinrespParDist.gnls
##alias<< kinrespParDist
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
##details<<
## Both the lognormal and the logitnormal distributions are skewed.
## Therefore, the mode, the median, and the expected value differ.
## At transformed, i.e. normal scale, they coincide. Simple backtransformation
## yield the median at original scale.
#mu0 <- coef(model)
mu0 <- fixef(model) #works on both gnls and nlme
if (!all(c("mumaxl","r0l","x0l") %in% names(mu0)) )
stop("kinrespParDist.gnls: provided model with some normal scale estimates missing.")
sigma0 <- summary(model)$tTable[,"Std.Error"] #same as sqrt(diag(model$varBeta)) for gls
sigma02 <- sigma0^2
cf0 <- confint(model)
# first assume lognormal distribution
colNames <- c("mean","sd","mle","median","cf025","cf975","mu","sigma")
res <- matrix(as.numeric(NA), nrow=3, ncol=length(colNames), dimnames=list( c("x0","r0","mumax"), colNames ))
#log-normal
i0 <- match(c("x0l","mumaxl"), names(mu0) )
mu <- mu0[i0]; sigma <- sigma0[i0]; sigma2 <- sigma02[i0]; cf<-cf0[i0,]
i <- match(c("x0","mumax"), rownames(res) )
res[i,"mean"] <- exp(mu+sigma2/2)
res[i,"sd"] <- sqrt( (exp(sigma2)-1)*exp( 2*mu+sigma2 ) )
res[i,"mle"] <- exp(mu-sigma2)
res[i,"median"] <- exp(mu)
res[i,c("cf025","cf975")] <- exp(cf)
res[i,c("mu","sigma")] <- cbind(mu,sigma)
#logit-normal
i0 <- match(c("r0l"), names(mu0) )
mu <- mu0[i0]; sigma <- sigma0[i0]; sigma2 <- sigma02[i0]; cf<-cf0[i0,]
i <- match(c("r0"), rownames(res) )
moments <- momentsLogitnorm( mu, sigma)
res[i,"mean"] <- moments["mean"]
res[i,"sd"] <- sqrt(moments["var"])
#mtrace(modeLogitnorm)
res[i,"mle"] <- modeLogitnorm(mu,sigma)
res[i,"median"] <- invlogit(mu)
res[i,c("cf025","cf975")] <- invlogit(cf)
res[i,c("mu","sigma")] <- cbind(mu,sigma)
res
### numeric matrix with rows corresponding to variables and columns \itemize{
### \item{ \code{mean}: expected value}
### \item{ \code{sd}: standard deviation}
### \item{ \code{mle}: the mode, i.e. the maximum likelihood estimate}
### \item{ \code{median}: the median}
### \item{ \code{cf025} and \code{cf975}: 2.5% and 97.5% confidence bounds}
### \item{ \code{mu} and \code{sigma}: parameters at normal, i.e log or logit transformed scale}
### }
})
setMethodS3("kinrespParDist","nlme", function(
### Expected value and confidence interval of microbial parameters
model ##<< the result of the \code{\link{fitKinrespReplicate}} or \code{\link{fitKinrespExperiment}$model}
,... ##<< currently not used
){
##seealso<<
## \code{\link{kinrespParDist.gnls}}
## ,\code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
kinrespParDist.gnls(model,...)
})
setMethodS3("coefKinresp","default", function(
### Check the microbial coefficients and translate to original microbial scale.
tmp.coef ##<< coefficients of model fit \code{coef(model)} (see details)
,...
){
# coefKinresp.default
##alias<< coefKinresp
##seealso<<
## \code{\link{twKinresp}}
##details<< \describe{\item{Functions for acccessing microbial parameters, and their uncertainty bounds.}{
## \itemize{
## \item{ Mode, Median, Mean and confidence bounds of microbial fits: \code{\link{kinrespParDist.gnls}}}
## \item{ Microbial parameters of single fit: this method (obtained by \code{coef(\link{kinrespGrowthphaseReplicate})} or \code{fixef(\link{fitKinrespExperiment})}) }
## \item{ Microbial parameters of replicate fits: \code{\link{coefKinrespMatrix}} (obtained by \code{coef(\link{fitKinrespExperiment})}) }
## \item{ Microbial parameters of a list of fits: \code{\link{coefKinresp.kinrespList}} (obtained by \code{\link{kinrespGrowthphaseExperiment}})}
## \item{ 95% confidenc interval of microbial parameters of a single fit: \code{\link{confintKinresp}} }
## \item{ Microbial parameters from beta-form of model fit: \code{\link{calcKinrespCoef}} }
## }
##}}
##details<< \describe{\item{Functions for translating between normal and original scale.}{
## \itemize{
## \item{ normalized to original scale: all the methods above. }
## \item{ original scale to normalized scale: \code{\link{coefKinrespNormStart}} }
## }
##}}
##details<< \describe{\item{ \code{coef(model)} }{
## Models are fitted in various forms differing by used coefficients.
## This method recognizes and translates coefficients of the following forms
## \itemize{
## \item{ beta-form at original scale and at normalized scale}
## \item{ beta-form excluding beta0 (fixed to 0 see
## \code{\link{calcKinrespCoef}}) }
## \item{ microbial form fit at original and transformed scale.}
## \item{ microbial form fit with excluding r0 (assuming r0=1)}
## }
## }}
if (names(tmp.coef)[1] %in% c("beta0l","beta0","beta1")) {
tmp.coef <- calcKinrespCoef(tmp.coef)
}
if ("x0l" %in% names(tmp.coef)) {
tmp.names <- names(tmp.coef)
tmp.names[match( "x0l", names(tmp.coef))] <- "x0"
names(tmp.coef) <- tmp.names
tmp.coef["x0"] <- exp(tmp.coef["x0"])
}
if ("r0l" %in% names(tmp.coef)){
tmp.names <- names(tmp.coef)
tmp.names[match( "r0l", names(tmp.coef))] <- "r0"
names(tmp.coef) <- tmp.names
tmp.coef["r0"] <- invlogit(tmp.coef["r0"])
}
if ("mumaxl" %in% names(tmp.coef)){
tmp.names <- names(tmp.coef)
tmp.names[match( "mumaxl", names(tmp.coef))] <- "mumax"
names(tmp.coef) <- tmp.names
tmp.coef["mumax"] <- exp(tmp.coef["mumax"])
}
if (!("r0" %in% names(tmp.coef)) ){
tmp.r0 = structure( attr(tmp.coef,"r0"), names=NULL )
if (is.null(tmp.r0) ) tmp.r0 = 1
tmp.coef <- c(tmp.coef["mumax"],r0=tmp.r0,tmp.coef["x0"])
}
tmp.coef
### named numeric vector (x0,r0,mumax)
})
setMethodS3("coefKinresp","kinrespList", function(
### Check the microbial coefficients and translate to original microbial scale for all replicates.
tmp.coef ##<< result of \code{\link{kinrespGrowthphaseExperiment}}
,rds.e=NULL ##<< constrained dataset, which may omit some replicates
,...
){
# coefKinresp.kinrespList
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
#resRepI <- tmp.coef$resRep[[1]]
bo <- if (is.null(rds.e)) TRUE else{
serUnique <- unique(getSERId(rds.e))
names(tmp.coef$resRep) %in% serUnique
}
tmp <- lapply( tmp.coef$resRep[bo], function(resRepI){ as.data.frame(c(list( experiment=resRepI$dataset$experiment[1], replicate=resRepI$dataset$replicate[1]), coefKinresp.default(coef(resRepI$fit)) ))})
do.call("rbind",tmp)
### named numer matrix (columns experiment, replicate, mumax, x0, and r0) with rows corresponding replicates
})
setMethodS3("fixef","kinresp", function(
### Make fixed.effects deliver attribute r0 if this is supplied with model.
object
,...
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
attr(object,"class") <- class(object)[-1]
tmp <- fixef(object)
if (!is.null(attr(object,"r0")) )
attr(tmp,"r0") <- attr(object,"r0")
tmp
})
setMethodS3("coef","kinresp", function(
### Make coef deliver attribute r0 if this is supplied with model.
object
,...
){
# coef.kinresp
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
attr(object,"class") <- class(object)[-1]
tmp <- coef(object)
if (!is.null(attr(object,"r0")) )
attr(tmp,"r0") <- attr(object,"r0")
tmp
})
setMethodS3("confint","kinresp", function(
### Make confint deliver attribute r0 if this was supplied with object
object
,...
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
attr(object,"class") <- class(object)[-1]
tmp <- confint(object)
if (!is.null(attr(object,"r0")) )
attr(tmp,"r0") <- attr(object,"r0")
tmp
})
coefKinrespMatrix <- function(
### Check the microbial coefficients and translates to original microbial scale.
tmp.coef ##<< multiple rows of microbial coefficients (see \code{\link{coefKinresp.default}})
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
if (colnames(tmp.coef)[1] %in% c("beta0l","beta0","beta1") ){
tmp.coef <- calcKinrespCoef(tmp.coef)
}
if ("x0l" %in% colnames(tmp.coef)){
tmp.colnames <- colnames(tmp.coef)
tmp.colnames[match( "x0l", colnames(tmp.coef))] <- "x0"
colnames(tmp.coef) <- tmp.colnames
tmp.coef[,"x0"] <- exp(tmp.coef[,"x0"])
}
if ("r0l" %in% colnames(tmp.coef)){
tmp.colnames <- colnames(tmp.coef)
tmp.colnames[match( "r0l", colnames(tmp.coef))] <- "r0"
colnames(tmp.coef) <- tmp.colnames
tmp.coef[,"r0"] <- invlogit(tmp.coef[,"r0"])
}
if ("mumaxl" %in% colnames(tmp.coef)){
tmp.colnames <- colnames(tmp.coef)
tmp.colnames[match( "mumaxl", colnames(tmp.coef))] <- "mumax"
colnames(tmp.coef) <- tmp.colnames
tmp.coef[,"mumax"] <- exp(tmp.coef[,"mumax"])
}
if (!("r0" %in% colnames(tmp.coef)) ){
tmp.r0 = structure( attr(tmp.coef,"r0"), colnames=NULL )
if (is.null(tmp.r0) ) tmp.r0 = 1
tmp.coef <- c(tmp.coef["mumax"],r0=tmp.r0,tmp.coef["x0"])
}
tmp.coef
}
confintKinresp <- function(
### Translate the confidence interval of estimated coefficients to original scale.
tmp.cf ##<< confidence interval from fitting the microbial model \code{confint(modelfit)}
){
# confintKinresp
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
tmp.i <- match("x0l", rownames(tmp.cf))
if (!is.na(tmp.i)){
tmp.cf[tmp.i,] <- exp(tmp.cf[tmp.i,])
rownames(tmp.cf)[tmp.i] <- "x0"
}
tmp.i <- match("r0l", rownames(tmp.cf))
if (!is.na(tmp.i)){
tmp.cf[tmp.i,] <- invlogit(tmp.cf[tmp.i,])
rownames(tmp.cf)[tmp.i] <- "r0"
}
tmp.i <- match("mumaxl", rownames(tmp.cf))
if (!is.na(tmp.i)){
tmp.cf[tmp.i,] <- exp(tmp.cf[tmp.i,])
rownames(tmp.cf)[tmp.i] <- "mumax"
}
if (rownames(tmp.cf)[2] != "r0" ){
tmp.r0 = structure( attr(tmp.cf,"r0"), names=NULL )
if (is.null(tmp.r0) ) tmp.r0 = 1
tmp.cf <- rbind( tmp.cf["mumax",],c( tmp.r0,tmp.r0), tmp.cf["x0",] )
rownames(tmp.cf) <- c("mumax","r0","x0")
}
tmp.cf
}
.coefKinrespLogStart <- function(
### Transform microbial coefficients to log scale (does ot care for r0)
tmp.coef
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
##seealso<< \code{\link{coefKinrespNormStart}}
tmp.coef <- coefKinresp(tmp.coef) # coefficients at the original scale including x0
tmp.names <- names(tmp.coef)
tmp.names[match( "x0", names(tmp.coef))] <- "x0l"
tmp.names[match( "mumax", names(tmp.coef))] <- "mumaxl"
names(tmp.coef) <- tmp.names
tmp.coef["x0l"] <- log(tmp.coef["x0l"])
tmp.coef["mumaxl"] <- log(tmp.coef["mumaxl"])
tmp.coef
}
coefKinrespNormStart <- function(
### Transform microbial coefficients to normal scale.
tmp.coef ##<< starting parameters of kinetic respiration, supplied to \code{\link{coefKinresp.default}} (maybe beta)
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
# coefKinrespNormStart
##details<<
## Replaces \itemize{
## \item{x0 by x0l = log(x0)}
## \item{mumax by mumaxl=log(mumax)} and
## \item{r0 by r0l=logit(r0l)}
## }
## Values at normalized scale are used as starting values for model fits.
tmp.coef <- coefKinresp(tmp.coef)
tmp.names <- names(tmp.coef)
tmp.names[match( "x0", names(tmp.coef))] <- "x0l"
tmp.names[match( "r0", names(tmp.coef))] <- "r0l"
tmp.names[match( "mumax", names(tmp.coef))] <- "mumaxl"
names(tmp.coef) <- tmp.names
tmp.coef["x0l"] <- log(tmp.coef["x0l"])
tmp.coef["r0l"] <- logit(tmp.coef["r0l"])
tmp.coef["mumaxl"] <- log(tmp.coef["mumaxl"])
tmp.coef
### named vector of coefficients (x0l,r0l,mumaxl)
}
Try the twKinresp package in your browser
Any scripts or data that you put into this service are public.
twKinresp documentation built on May 2, 2019, 4:47 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions coefKinrespMatrix confintKinresp .coefKinrespLogStart coefKinrespNormStart
Documented in coefKinrespMatrix coefKinrespNormStart confintKinresp
setMethodS3("kinrespParDist","gnls", function(
### Expected value and confidence interval of microbial parameters
model ##<< the result of the \code{\link{fitKinrespReplicate}} or \code{\link{fitKinrespExperiment}$model}
,... ##<< currently not used
){
# kinrespParDist.gnls
##alias<< kinrespParDist
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
##details<<
## Both the lognormal and the logitnormal distributions are skewed.
## Therefore, the mode, the median, and the expected value differ.
## At transformed, i.e. normal scale, they coincide. Simple backtransformation
## yield the median at original scale.
#mu0 <- coef(model)
mu0 <- fixef(model) #works on both gnls and nlme
if (!all(c("mumaxl","r0l","x0l") %in% names(mu0)) )
stop("kinrespParDist.gnls: provided model with some normal scale estimates missing.")
sigma0 <- summary(model)$tTable[,"Std.Error"] #same as sqrt(diag(model$varBeta)) for gls
sigma02 <- sigma0^2
cf0 <- confint(model)
# first assume lognormal distribution
colNames <- c("mean","sd","mle","median","cf025","cf975","mu","sigma")
res <- matrix(as.numeric(NA), nrow=3, ncol=length(colNames), dimnames=list( c("x0","r0","mumax"), colNames ))
#log-normal
i0 <- match(c("x0l","mumaxl"), names(mu0) )
mu <- mu0[i0]; sigma <- sigma0[i0]; sigma2 <- sigma02[i0]; cf<-cf0[i0,]
i <- match(c("x0","mumax"), rownames(res) )
res[i,"mean"] <- exp(mu+sigma2/2)
res[i,"sd"] <- sqrt( (exp(sigma2)-1)*exp( 2*mu+sigma2 ) )
res[i,"mle"] <- exp(mu-sigma2)
res[i,"median"] <- exp(mu)
res[i,c("cf025","cf975")] <- exp(cf)
res[i,c("mu","sigma")] <- cbind(mu,sigma)
#logit-normal
i0 <- match(c("r0l"), names(mu0) )
mu <- mu0[i0]; sigma <- sigma0[i0]; sigma2 <- sigma02[i0]; cf<-cf0[i0,]
i <- match(c("r0"), rownames(res) )
moments <- momentsLogitnorm( mu, sigma)
res[i,"mean"] <- moments["mean"]
res[i,"sd"] <- sqrt(moments["var"])
#mtrace(modeLogitnorm)
res[i,"mle"] <- modeLogitnorm(mu,sigma)
res[i,"median"] <- invlogit(mu)
res[i,c("cf025","cf975")] <- invlogit(cf)
res[i,c("mu","sigma")] <- cbind(mu,sigma)
res
### numeric matrix with rows corresponding to variables and columns \itemize{
### \item{ \code{mean}: expected value}
### \item{ \code{sd}: standard deviation}
### \item{ \code{mle}: the mode, i.e. the maximum likelihood estimate}
### \item{ \code{median}: the median}
### \item{ \code{cf025} and \code{cf975}: 2.5% and 97.5% confidence bounds}
### \item{ \code{mu} and \code{sigma}: parameters at normal, i.e log or logit transformed scale}
### }
})
setMethodS3("kinrespParDist","nlme", function(
### Expected value and confidence interval of microbial parameters
model ##<< the result of the \code{\link{fitKinrespReplicate}} or \code{\link{fitKinrespExperiment}$model}
,... ##<< currently not used
){
##seealso<<
## \code{\link{kinrespParDist.gnls}}
## ,\code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
kinrespParDist.gnls(model,...)
})
setMethodS3("coefKinresp","default", function(
### Check the microbial coefficients and translate to original microbial scale.
tmp.coef ##<< coefficients of model fit \code{coef(model)} (see details)
,...
){
# coefKinresp.default
##alias<< coefKinresp
##seealso<<
## \code{\link{twKinresp}}
##details<< \describe{\item{Functions for acccessing microbial parameters, and their uncertainty bounds.}{
## \itemize{
## \item{ Mode, Median, Mean and confidence bounds of microbial fits: \code{\link{kinrespParDist.gnls}}}
## \item{ Microbial parameters of single fit: this method (obtained by \code{coef(\link{kinrespGrowthphaseReplicate})} or \code{fixef(\link{fitKinrespExperiment})}) }
## \item{ Microbial parameters of replicate fits: \code{\link{coefKinrespMatrix}} (obtained by \code{coef(\link{fitKinrespExperiment})}) }
## \item{ Microbial parameters of a list of fits: \code{\link{coefKinresp.kinrespList}} (obtained by \code{\link{kinrespGrowthphaseExperiment}})}
## \item{ 95% confidenc interval of microbial parameters of a single fit: \code{\link{confintKinresp}} }
## \item{ Microbial parameters from beta-form of model fit: \code{\link{calcKinrespCoef}} }
## }
##}}
##details<< \describe{\item{Functions for translating between normal and original scale.}{
## \itemize{
## \item{ normalized to original scale: all the methods above. }
## \item{ original scale to normalized scale: \code{\link{coefKinrespNormStart}} }
## }
##}}
##details<< \describe{\item{ \code{coef(model)} }{
## Models are fitted in various forms differing by used coefficients.
## This method recognizes and translates coefficients of the following forms
## \itemize{
## \item{ beta-form at original scale and at normalized scale}
## \item{ beta-form excluding beta0 (fixed to 0 see
## \code{\link{calcKinrespCoef}}) }
## \item{ microbial form fit at original and transformed scale.}
## \item{ microbial form fit with excluding r0 (assuming r0=1)}
## }
## }}
if (names(tmp.coef)[1] %in% c("beta0l","beta0","beta1")) {
tmp.coef <- calcKinrespCoef(tmp.coef)
}
if ("x0l" %in% names(tmp.coef)) {
tmp.names <- names(tmp.coef)
tmp.names[match( "x0l", names(tmp.coef))] <- "x0"
names(tmp.coef) <- tmp.names
tmp.coef["x0"] <- exp(tmp.coef["x0"])
}
if ("r0l" %in% names(tmp.coef)){
tmp.names <- names(tmp.coef)
tmp.names[match( "r0l", names(tmp.coef))] <- "r0"
names(tmp.coef) <- tmp.names
tmp.coef["r0"] <- invlogit(tmp.coef["r0"])
}
if ("mumaxl" %in% names(tmp.coef)){
tmp.names <- names(tmp.coef)
tmp.names[match( "mumaxl", names(tmp.coef))] <- "mumax"
names(tmp.coef) <- tmp.names
tmp.coef["mumax"] <- exp(tmp.coef["mumax"])
}
if (!("r0" %in% names(tmp.coef)) ){
tmp.r0 = structure( attr(tmp.coef,"r0"), names=NULL )
if (is.null(tmp.r0) ) tmp.r0 = 1
tmp.coef <- c(tmp.coef["mumax"],r0=tmp.r0,tmp.coef["x0"])
}
tmp.coef
### named numeric vector (x0,r0,mumax)
})
setMethodS3("coefKinresp","kinrespList", function(
### Check the microbial coefficients and translate to original microbial scale for all replicates.
tmp.coef ##<< result of \code{\link{kinrespGrowthphaseExperiment}}
,rds.e=NULL ##<< constrained dataset, which may omit some replicates
,...
){
# coefKinresp.kinrespList
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
#resRepI <- tmp.coef$resRep[[1]]
bo <- if (is.null(rds.e)) TRUE else{
serUnique <- unique(getSERId(rds.e))
names(tmp.coef$resRep) %in% serUnique
}
tmp <- lapply( tmp.coef$resRep[bo], function(resRepI){ as.data.frame(c(list( experiment=resRepI$dataset$experiment[1], replicate=resRepI$dataset$replicate[1]), coefKinresp.default(coef(resRepI$fit)) ))})
do.call("rbind",tmp)
### named numer matrix (columns experiment, replicate, mumax, x0, and r0) with rows corresponding replicates
})
setMethodS3("fixef","kinresp", function(
### Make fixed.effects deliver attribute r0 if this is supplied with model.
object
,...
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
attr(object,"class") <- class(object)[-1]
tmp <- fixef(object)
if (!is.null(attr(object,"r0")) )
attr(tmp,"r0") <- attr(object,"r0")
tmp
})
setMethodS3("coef","kinresp", function(
### Make coef deliver attribute r0 if this is supplied with model.
object
,...
){
# coef.kinresp
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
attr(object,"class") <- class(object)[-1]
tmp <- coef(object)
if (!is.null(attr(object,"r0")) )
attr(tmp,"r0") <- attr(object,"r0")
tmp
})
setMethodS3("confint","kinresp", function(
### Make confint deliver attribute r0 if this was supplied with object
object
,...
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
attr(object,"class") <- class(object)[-1]
tmp <- confint(object)
if (!is.null(attr(object,"r0")) )
attr(tmp,"r0") <- attr(object,"r0")
tmp
})
coefKinrespMatrix <- function(
### Check the microbial coefficients and translates to original microbial scale.
tmp.coef ##<< multiple rows of microbial coefficients (see \code{\link{coefKinresp.default}})
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
if (colnames(tmp.coef)[1] %in% c("beta0l","beta0","beta1") ){
tmp.coef <- calcKinrespCoef(tmp.coef)
}
if ("x0l" %in% colnames(tmp.coef)){
tmp.colnames <- colnames(tmp.coef)
tmp.colnames[match( "x0l", colnames(tmp.coef))] <- "x0"
colnames(tmp.coef) <- tmp.colnames
tmp.coef[,"x0"] <- exp(tmp.coef[,"x0"])
}
if ("r0l" %in% colnames(tmp.coef)){
tmp.colnames <- colnames(tmp.coef)
tmp.colnames[match( "r0l", colnames(tmp.coef))] <- "r0"
colnames(tmp.coef) <- tmp.colnames
tmp.coef[,"r0"] <- invlogit(tmp.coef[,"r0"])
}
if ("mumaxl" %in% colnames(tmp.coef)){
tmp.colnames <- colnames(tmp.coef)
tmp.colnames[match( "mumaxl", colnames(tmp.coef))] <- "mumax"
colnames(tmp.coef) <- tmp.colnames
tmp.coef[,"mumax"] <- exp(tmp.coef[,"mumax"])
}
if (!("r0" %in% colnames(tmp.coef)) ){
tmp.r0 = structure( attr(tmp.coef,"r0"), colnames=NULL )
if (is.null(tmp.r0) ) tmp.r0 = 1
tmp.coef <- c(tmp.coef["mumax"],r0=tmp.r0,tmp.coef["x0"])
}
tmp.coef
}
confintKinresp <- function(
### Translate the confidence interval of estimated coefficients to original scale.
tmp.cf ##<< confidence interval from fitting the microbial model \code{confint(modelfit)}
){
# confintKinresp
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
tmp.i <- match("x0l", rownames(tmp.cf))
if (!is.na(tmp.i)){
tmp.cf[tmp.i,] <- exp(tmp.cf[tmp.i,])
rownames(tmp.cf)[tmp.i] <- "x0"
}
tmp.i <- match("r0l", rownames(tmp.cf))
if (!is.na(tmp.i)){
tmp.cf[tmp.i,] <- invlogit(tmp.cf[tmp.i,])
rownames(tmp.cf)[tmp.i] <- "r0"
}
tmp.i <- match("mumaxl", rownames(tmp.cf))
if (!is.na(tmp.i)){
tmp.cf[tmp.i,] <- exp(tmp.cf[tmp.i,])
rownames(tmp.cf)[tmp.i] <- "mumax"
}
if (rownames(tmp.cf)[2] != "r0" ){
tmp.r0 = structure( attr(tmp.cf,"r0"), names=NULL )
if (is.null(tmp.r0) ) tmp.r0 = 1
tmp.cf <- rbind( tmp.cf["mumax",],c( tmp.r0,tmp.r0), tmp.cf["x0",] )
rownames(tmp.cf) <- c("mumax","r0","x0")
}
tmp.cf
}
.coefKinrespLogStart <- function(
### Transform microbial coefficients to log scale (does ot care for r0)
tmp.coef
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
##seealso<< \code{\link{coefKinrespNormStart}}
tmp.coef <- coefKinresp(tmp.coef) # coefficients at the original scale including x0
tmp.names <- names(tmp.coef)
tmp.names[match( "x0", names(tmp.coef))] <- "x0l"
tmp.names[match( "mumax", names(tmp.coef))] <- "mumaxl"
names(tmp.coef) <- tmp.names
tmp.coef["x0l"] <- log(tmp.coef["x0l"])
tmp.coef["mumaxl"] <- log(tmp.coef["mumaxl"])
tmp.coef
}
coefKinrespNormStart <- function(
### Transform microbial coefficients to normal scale.
tmp.coef ##<< starting parameters of kinetic respiration, supplied to \code{\link{coefKinresp.default}} (maybe beta)
){
##seealso<<
## \code{\link{coefKinresp.default}}
## ,\code{\link{twKinresp}}
# coefKinrespNormStart
##details<<
## Replaces \itemize{
## \item{x0 by x0l = log(x0)}
## \item{mumax by mumaxl=log(mumax)} and
## \item{r0 by r0l=logit(r0l)}
## }
## Values at normalized scale are used as starting values for model fits.
tmp.coef <- coefKinresp(tmp.coef)
tmp.names <- names(tmp.coef)
tmp.names[match( "x0", names(tmp.coef))] <- "x0l"
tmp.names[match( "r0", names(tmp.coef))] <- "r0l"
tmp.names[match( "mumax", names(tmp.coef))] <- "mumaxl"
names(tmp.coef) <- tmp.names
tmp.coef["x0l"] <- log(tmp.coef["x0l"])
tmp.coef["r0l"] <- logit(tmp.coef["r0l"])
tmp.coef["mumaxl"] <- log(tmp.coef["mumaxl"])
tmp.coef
### named vector of coefficients (x0l,r0l,mumaxl)
}
Try the twKinresp package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.