R/confint_methods.R
In AndreasChristianHill/forestinventory: Design-Based Global and Small-Area Estimations for Multiphase Forest Inventories
Defines functions
confint.threephase
confint.twophase
confint.onephase
Documented in
confint.onephase
confint.threephase
confint.twophase
#' Calculates Confidence Intervals for Global and Small-Area Estimations
#'
#' @param object object of class \code{onephase}, \code{twophase} or \code{threephase},
#' containing estimation results of the respective estimation method.
#' @param parm ignored.
#' @param level the confidence level required.
#' @param ... additional arguments, so far ignored.
#' @param adjust.method correction method to obtain simultaneous confidence intervals
#' for a set of estimates (thus restricted to objects of class \code{"onephase"},
#' \code{c("smallarea", "twophase")} and \code{c("smallarea", "threephase")}).
#' Available correction methods are \code{c("none","bonferroni")}. Defaults to \code{"none"}.
#'
#' @details
#'
#' Depending on the estimation method specified, \code{confint()} computes confidence intervals as follows:
#'
#' \code{\link{onephase}}:
#'
#' Two-sided confidence intervals are computed based on the t-distribution with \code{n2 - 1} \emph{degrees of freedom},
#' where \code{n2} is the number of terrestrial data in the respective inventory domain.
#'
#' \code{\link{twophase}}:
#'
#' The calculation of the two-sided confidence intervals for \emph{global} twophase estimates
#' (objects of class \code{global}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2 - p} \emph{degrees of freedom}, where \code{p} is the number of parameters used in
#' the regression model, and \code{n2} is the number of terrestrial observations (i.e. \emph{local densities})
#' in the inventory domain.
#'
#' The calculation of the two-sided confidence intervals for \emph{smallarea} twophase estimates
#' (objects of class \code{smallarea}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2G - 1} \emph{degrees of freedom}, where \code{n2G} is the number of
#' terrestrial observations (i.e. \emph{local densities}) in the smallarea.
#'
#' \code{\link{threephase}}:
#'
#' The calculation of the two-sided confidence intervals for \emph{global} threephase estimates
#' (objects of class \code{global}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2 - p} \emph{degrees of freedom}, where \code{p} is the number of parameters used in
#' the \strong{full} regression model, and \code{n2} is the number of terrestrial observations
#' (i.e. \emph{local densities}) in the inventory domain (note: in notation used here n0, n1 and n2
#' correspond to the zero, first and second phase sample sizes respectively).
#'
#' The calculation of the two-sided confidence intervals for \emph{smallarea} theephase estimates
#' (objects of class \code{smallarea}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2G - 1} \emph{degrees of freedom}, where \code{n2G} is the number of
#' terrestrial observations (i.e. \emph{local densities}) in the smallarea.
#'
#'@return \code{confint} returns a list of the following 3 components:
#'
#' \item{ci}{a \code{data.frame} containing the columns:
#' \itemize{
#' \item \code{area} the domain, i.e. small area
#' \item \code{ci_lower_ext} the lower confidence limit based on the external variance
#' \item \code{ci_upper_ext} the upper confidence limit based on the external variance
#' \item \code{ci_lower_g} the lower confidence limit based on the g-weight variance
#' \item \code{ci_upper_g} the upper confidence limit based on the g-weight variance
#' }}
#' \item{level}{the applied confidence level}
#' \item{adjust.method}{the adjustment method applied to retrieve simultaneous confidence intervals}
#'
#'
#'
#' @note
#' In the special case of \emph{synthetic} smallarea estimations, the two-sided confidence intervals
#' are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2 - p} \emph{degrees of freedom}, i.e. based on the global sample size.
#'
#' The confidence intervals for \emph{synthetic} smallarea estimations do not account for the potential
#' bias of a linear model that was fit in a large forest area and applied to a small area. Thus,
#' the coverage rates for confidence intervals produced by synthetic estimators may be less than
#' the nominal level of confidence.
#'
#' In case of cluster-sampling, \code{n2G} is the number of terrestrial clusters
#' (a cluster constitutes the sample unit). This is automatically considered by \code{confint}.
#'
#' The adjustment methods passed to \code{adjust.method} are designed to achieve
#' \emph{simultaneous} confidence intervals by correcting the confidence level given by \code{level}.
#' The use of this option is recommended if a set of estimates contained in a \code{onephase}- or
#' \code{smallarea}-object should be compared by their confidence intervals. It ensures that the
#' percentage of confidence intervals containing the true value will correspond to
#' the nominal confidence level.
#'
#'
#' @references
#'
#' Hill, A., Massey, A. F. (2021). \emph{The R Package forestinventory: Design-Based Global and Small
#' Area Estimations for Multiphase Forest Inventories.} Journal of Statistical Software, 97(4), 1-40.
#'
#' Mandallaz, D. (2013). \emph{Design-based properties of some small-area estimators in forest inventory
#' with two-phase sampling.} Canadian Journal of Forest Research, 43(5), 441-449.
#'
#' Mandallaz, D., Breschan, J., & Hill, A. (2013). \emph{New regression estimators in forest inventories
#' with two-phase sampling and partially exhaustive information: a design-based monte carlo approach
#' with applications to small-area estimation.} Canadian Journal of Forest Research,
#' 43(11), 1023-1031.
#'
#' Mandallaz, D. (2013). \emph{A three-phase sampling extension of the generalized regression estimator
#' with partially exhaustive information.} Canadian Journal of Forest Research,
#' 44(4), 383-388.
#'
#' Benjamini, Y., and Hochberg, Y. (1995). \emph{Controlling the false discovery rate:
#' a practical and powerful approach to multiple testing.}
#' Journal of the Royal Statistical Society Series B 57, 289-300.
#'
#' @example examples/example_confint.R
#' @import methods
#' @name confint
NULL
#>
# -----------------------------------------------------------------------------#
# -----------------------------------------------------------------------------#
#' @rdname confint
#' @export
confint.onephase<- function(object, parm, level = 0.95, adjust.method="none",...){
# confint-method for one-phase outputs:
orig.level<- level
# ----------#
# adjust confidence level if required:
if(adjust.method %in% c("none", "bonferroni")){
# bonferroni-correction:
if (adjust.method == "bonferroni"){
if(nrow(object$estimation)==1){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
} else {
level<- 1 - ((1 - level) / nrow(object$estimation))
}
}
} else {
stop("invalid argument for 'adjust.method'")
}
# ----------#
# calculating the quantile of the t-distribution with n2 - 1 df:
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = object$estimation$n2 - 1))
# calculating the lower and upper confidence intervals based on the design-based variance:
ci_lower_op<- object$estimation$estimate - (t.val * sqrt(object$estimation$variance))
ci_upper_op<- object$estimation$estimate + (t.val * sqrt(object$estimation$variance))
if ("area" %in% names(object$estimation)){
ci<- list(ci=data.frame(area=object$estimation$area, estimate=object$estimation$estimate, ci_lower_op, ci_upper_op), level=orig.level, adjust.method=adjust.method)
} else {
ci<- list(ci=data.frame(estimate=object$estimation$estimate, ci_lower_op, ci_upper_op), level=orig.level, adjust.method=adjust.method)
}
class(ci)<- c("confint.global", "onephase")
return(ci)
} # end of confint.onephase
# -----------------------------------------------------------------------------#
# -----------------------------------------------------------------------------#
#' @rdname confint
#' @export
confint.twophase<- function(object, parm, level = 0.95, adjust.method="none",...){
# confint-method for for twophase estimations:
orig.level<- level
# -------------------------------------------#
# confidence-intervals for twophase-smallarea:
if(is(object,"smallarea") & inherits(object, "twophase")){ # if class(object) is c("smallarea", "twophase")
# ----------#
# adjust confidence level if required:
if(adjust.method %in% c("none", "bonferroni")){
# bonferroni-correction:
if (adjust.method == "bonferroni"){
if(nrow(object$estimation)==1){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
} else {
level<- 1 - ((1 - level) / nrow(object$estimation))
}
}
} else {
stop("invalid argument for 'adjust.method'")
}
# ----------#
# calulating the quantile of the t-distribution:
# if synthetic-estimation was applied, the df's are calculated as n2 - # parameters of regression model:
if(object$input$method %in% c("synth", "psynth")){
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = unique(object$estimation[["n2"]]) - length(all.vars(object$input$formula)[-1])))
# if non-synthetic-estimation was applied, the df's are calculated as n2G - 1:
} else {
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = object$estimation$n2G - 1))
}
#-----------#
# calculating the lower and upper confidence intervals based on the external variance:
ci_lower_g<- object$estimation$estimate - (t.val * sqrt(object$estimation$g_variance))
ci_upper_g<- object$estimation$estimate + (t.val * sqrt(object$estimation$g_variance))
# calculating the lower and upper confidence intervals based on the design-based variance:
ci_lower_ext<- object$estimation$estimate - (t.val * sqrt(object$estimation$ext_variance))
ci_upper_ext<- object$estimation$estimate + (t.val * sqrt(object$estimation$ext_variance))
ci<- list(ci=data.frame(area=object$estimation$area, estimate=object$estimation$estimate, ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level, adjust.method=adjust.method)
class(ci)<- c("confint.smallarea", "twophase")
return(ci)
} # end of smallarea-CI
# -------------------------------------------#
# confidence-intervals for twophase-global:
if(is(object, "global") & inherits(object, "twophase")){ # if class(object) is c("global", "twophase")
if(adjust.method != "none"){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
}
n2<- ifelse(object$input$cluster, object$samplesizes$n2_clust, object$samplesizes$n2)
t.val<- qt(p = 1-( (1-level)/2 ), df = n2 - length(all.vars(object$input$formula)[-1]))
ci_lower_g<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["g_variance"]]))
ci_upper_g<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["g_variance"]]))
ci_lower_ext<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["ext_variance"]]))
ci_upper_ext<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["ext_variance"]]))
ci<- list(ci=data.frame(estimate=object$estimation[["estimate"]], ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level, adjust.method=adjust.method)
class(ci)<- c("confint.global", "twophase")
return(ci)
}# end of global-CI
} ## end of confint.twophase
# -----------------------------------------------------------------------------#
# -----------------------------------------------------------------------------#
#' @rdname confint
#' @export
confint.threephase<- function(object, parm, level = 0.95, adjust.method="none", ...){
# confint-method for for threephase estimations:
orig.level<- level
# --------------------------------#
# confint for threephase-smallarea:
if(is(object, "smallarea") & inherits(object, "threephase")){ # if class(object) is c("smallarea", "threephase")
#-----------#
# adjust confidence level if required:
if(adjust.method %in% c("none", "bonferroni")){
# bonferroni-correction:
if (adjust.method == "bonferroni"){
if(nrow(object$estimation)==1){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
} else {
level<- 1 - ((1 - level) / nrow(object$estimation))
}
}
} else {
stop("invalid argument for 'adjust.method'")
}
#-----------#
# calulating the quantile of the t-distribution:
# if synthetic-estimation was applied, the df's are calculated as n2 - # parameters of regression model:
if(object$input$method %in% c("synth", "psynth")){
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = unique(object$estimation[["n2"]]) - length(all.vars(object$input$formula.s1)[-1])))
# if non-synthetic-estimation was applied, the df's are calculated as n2G - 1:
} else {
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = object$estimation$n2G - 1))
}
#-----------#
# calculating the lower and upper confidence intervals based on the external variance:
ci_lower_g<- object$estimation$estimate - (t.val * sqrt(object$estimation$g_variance))
ci_upper_g<- object$estimation$estimate + (t.val * sqrt(object$estimation$g_variance))
# calculating the lower and upper confidence intervals based on the design-based variance:
ci_lower_ext<- object$estimation$estimate - (t.val * sqrt(object$estimation$ext_variance))
ci_upper_ext<- object$estimation$estimate + (t.val * sqrt(object$estimation$ext_variance))
ci<- list(ci=data.frame(area=object$estimation$area, estimate=object$estimation$estimate, ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level, adjust.method=adjust.method)
class(ci)<- c("confint.smallarea", "threephase")
return(ci)
} # end of smallarea-CI
# --------------------------------#
# summary for threephase-global:
if(is(object, "global") & inherits(object, "threephase")){ # if class(object) is c("global", "threephase")
if(adjust.method != "none"){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
}
n2<- ifelse(object$input$cluster, object$samplesizes$n2_clust, object$samplesizes$n2)
t.val<- qt(p = 1-( (1-level)/2 ), df = n2 - length(all.vars(object$input$formula.s1)[-1])) # that's however not recommended by Daniel,
# see Article Crfr, Threephase, page 387
ci_lower_g<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["g_variance"]]))
ci_upper_g<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["g_variance"]]))
ci_lower_ext<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["ext_variance"]]))
ci_upper_ext<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["ext_variance"]]))
ci<- list(ci=data.frame(estimate=object$estimation[["estimate"]], ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level,adjust.method=adjust.method)
class(ci)<- c("confint.global", "threepase")
return(ci)
}# end of global-CI
}## end of confint.threephase
AndreasChristianHill/forestinventory documentation built on Aug. 16, 2021, 2:13 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 confint.threephase confint.twophase confint.onephase
Documented in confint.onephase confint.threephase confint.twophase
#' Calculates Confidence Intervals for Global and Small-Area Estimations
#'
#' @param object object of class \code{onephase}, \code{twophase} or \code{threephase},
#' containing estimation results of the respective estimation method.
#' @param parm ignored.
#' @param level the confidence level required.
#' @param ... additional arguments, so far ignored.
#' @param adjust.method correction method to obtain simultaneous confidence intervals
#' for a set of estimates (thus restricted to objects of class \code{"onephase"},
#' \code{c("smallarea", "twophase")} and \code{c("smallarea", "threephase")}).
#' Available correction methods are \code{c("none","bonferroni")}. Defaults to \code{"none"}.
#'
#' @details
#'
#' Depending on the estimation method specified, \code{confint()} computes confidence intervals as follows:
#'
#' \code{\link{onephase}}:
#'
#' Two-sided confidence intervals are computed based on the t-distribution with \code{n2 - 1} \emph{degrees of freedom},
#' where \code{n2} is the number of terrestrial data in the respective inventory domain.
#'
#' \code{\link{twophase}}:
#'
#' The calculation of the two-sided confidence intervals for \emph{global} twophase estimates
#' (objects of class \code{global}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2 - p} \emph{degrees of freedom}, where \code{p} is the number of parameters used in
#' the regression model, and \code{n2} is the number of terrestrial observations (i.e. \emph{local densities})
#' in the inventory domain.
#'
#' The calculation of the two-sided confidence intervals for \emph{smallarea} twophase estimates
#' (objects of class \code{smallarea}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2G - 1} \emph{degrees of freedom}, where \code{n2G} is the number of
#' terrestrial observations (i.e. \emph{local densities}) in the smallarea.
#'
#' \code{\link{threephase}}:
#'
#' The calculation of the two-sided confidence intervals for \emph{global} threephase estimates
#' (objects of class \code{global}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2 - p} \emph{degrees of freedom}, where \code{p} is the number of parameters used in
#' the \strong{full} regression model, and \code{n2} is the number of terrestrial observations
#' (i.e. \emph{local densities}) in the inventory domain (note: in notation used here n0, n1 and n2
#' correspond to the zero, first and second phase sample sizes respectively).
#'
#' The calculation of the two-sided confidence intervals for \emph{smallarea} theephase estimates
#' (objects of class \code{smallarea}) are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2G - 1} \emph{degrees of freedom}, where \code{n2G} is the number of
#' terrestrial observations (i.e. \emph{local densities}) in the smallarea.
#'
#'@return \code{confint} returns a list of the following 3 components:
#'
#' \item{ci}{a \code{data.frame} containing the columns:
#' \itemize{
#' \item \code{area} the domain, i.e. small area
#' \item \code{ci_lower_ext} the lower confidence limit based on the external variance
#' \item \code{ci_upper_ext} the upper confidence limit based on the external variance
#' \item \code{ci_lower_g} the lower confidence limit based on the g-weight variance
#' \item \code{ci_upper_g} the upper confidence limit based on the g-weight variance
#' }}
#' \item{level}{the applied confidence level}
#' \item{adjust.method}{the adjustment method applied to retrieve simultaneous confidence intervals}
#'
#'
#'
#' @note
#' In the special case of \emph{synthetic} smallarea estimations, the two-sided confidence intervals
#' are calculated based on the quantiles of the \emph{t}-distribution
#' with \code{n2 - p} \emph{degrees of freedom}, i.e. based on the global sample size.
#'
#' The confidence intervals for \emph{synthetic} smallarea estimations do not account for the potential
#' bias of a linear model that was fit in a large forest area and applied to a small area. Thus,
#' the coverage rates for confidence intervals produced by synthetic estimators may be less than
#' the nominal level of confidence.
#'
#' In case of cluster-sampling, \code{n2G} is the number of terrestrial clusters
#' (a cluster constitutes the sample unit). This is automatically considered by \code{confint}.
#'
#' The adjustment methods passed to \code{adjust.method} are designed to achieve
#' \emph{simultaneous} confidence intervals by correcting the confidence level given by \code{level}.
#' The use of this option is recommended if a set of estimates contained in a \code{onephase}- or
#' \code{smallarea}-object should be compared by their confidence intervals. It ensures that the
#' percentage of confidence intervals containing the true value will correspond to
#' the nominal confidence level.
#'
#'
#' @references
#'
#' Hill, A., Massey, A. F. (2021). \emph{The R Package forestinventory: Design-Based Global and Small
#' Area Estimations for Multiphase Forest Inventories.} Journal of Statistical Software, 97(4), 1-40.
#'
#' Mandallaz, D. (2013). \emph{Design-based properties of some small-area estimators in forest inventory
#' with two-phase sampling.} Canadian Journal of Forest Research, 43(5), 441-449.
#'
#' Mandallaz, D., Breschan, J., & Hill, A. (2013). \emph{New regression estimators in forest inventories
#' with two-phase sampling and partially exhaustive information: a design-based monte carlo approach
#' with applications to small-area estimation.} Canadian Journal of Forest Research,
#' 43(11), 1023-1031.
#'
#' Mandallaz, D. (2013). \emph{A three-phase sampling extension of the generalized regression estimator
#' with partially exhaustive information.} Canadian Journal of Forest Research,
#' 44(4), 383-388.
#'
#' Benjamini, Y., and Hochberg, Y. (1995). \emph{Controlling the false discovery rate:
#' a practical and powerful approach to multiple testing.}
#' Journal of the Royal Statistical Society Series B 57, 289-300.
#'
#' @example examples/example_confint.R
#' @import methods
#' @name confint
NULL
#>
# -----------------------------------------------------------------------------#
# -----------------------------------------------------------------------------#
#' @rdname confint
#' @export
confint.onephase<- function(object, parm, level = 0.95, adjust.method="none",...){
# confint-method for one-phase outputs:
orig.level<- level
# ----------#
# adjust confidence level if required:
if(adjust.method %in% c("none", "bonferroni")){
# bonferroni-correction:
if (adjust.method == "bonferroni"){
if(nrow(object$estimation)==1){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
} else {
level<- 1 - ((1 - level) / nrow(object$estimation))
}
}
} else {
stop("invalid argument for 'adjust.method'")
}
# ----------#
# calculating the quantile of the t-distribution with n2 - 1 df:
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = object$estimation$n2 - 1))
# calculating the lower and upper confidence intervals based on the design-based variance:
ci_lower_op<- object$estimation$estimate - (t.val * sqrt(object$estimation$variance))
ci_upper_op<- object$estimation$estimate + (t.val * sqrt(object$estimation$variance))
if ("area" %in% names(object$estimation)){
ci<- list(ci=data.frame(area=object$estimation$area, estimate=object$estimation$estimate, ci_lower_op, ci_upper_op), level=orig.level, adjust.method=adjust.method)
} else {
ci<- list(ci=data.frame(estimate=object$estimation$estimate, ci_lower_op, ci_upper_op), level=orig.level, adjust.method=adjust.method)
}
class(ci)<- c("confint.global", "onephase")
return(ci)
} # end of confint.onephase
# -----------------------------------------------------------------------------#
# -----------------------------------------------------------------------------#
#' @rdname confint
#' @export
confint.twophase<- function(object, parm, level = 0.95, adjust.method="none",...){
# confint-method for for twophase estimations:
orig.level<- level
# -------------------------------------------#
# confidence-intervals for twophase-smallarea:
if(is(object,"smallarea") & inherits(object, "twophase")){ # if class(object) is c("smallarea", "twophase")
# ----------#
# adjust confidence level if required:
if(adjust.method %in% c("none", "bonferroni")){
# bonferroni-correction:
if (adjust.method == "bonferroni"){
if(nrow(object$estimation)==1){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
} else {
level<- 1 - ((1 - level) / nrow(object$estimation))
}
}
} else {
stop("invalid argument for 'adjust.method'")
}
# ----------#
# calulating the quantile of the t-distribution:
# if synthetic-estimation was applied, the df's are calculated as n2 - # parameters of regression model:
if(object$input$method %in% c("synth", "psynth")){
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = unique(object$estimation[["n2"]]) - length(all.vars(object$input$formula)[-1])))
# if non-synthetic-estimation was applied, the df's are calculated as n2G - 1:
} else {
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = object$estimation$n2G - 1))
}
#-----------#
# calculating the lower and upper confidence intervals based on the external variance:
ci_lower_g<- object$estimation$estimate - (t.val * sqrt(object$estimation$g_variance))
ci_upper_g<- object$estimation$estimate + (t.val * sqrt(object$estimation$g_variance))
# calculating the lower and upper confidence intervals based on the design-based variance:
ci_lower_ext<- object$estimation$estimate - (t.val * sqrt(object$estimation$ext_variance))
ci_upper_ext<- object$estimation$estimate + (t.val * sqrt(object$estimation$ext_variance))
ci<- list(ci=data.frame(area=object$estimation$area, estimate=object$estimation$estimate, ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level, adjust.method=adjust.method)
class(ci)<- c("confint.smallarea", "twophase")
return(ci)
} # end of smallarea-CI
# -------------------------------------------#
# confidence-intervals for twophase-global:
if(is(object, "global") & inherits(object, "twophase")){ # if class(object) is c("global", "twophase")
if(adjust.method != "none"){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
}
n2<- ifelse(object$input$cluster, object$samplesizes$n2_clust, object$samplesizes$n2)
t.val<- qt(p = 1-( (1-level)/2 ), df = n2 - length(all.vars(object$input$formula)[-1]))
ci_lower_g<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["g_variance"]]))
ci_upper_g<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["g_variance"]]))
ci_lower_ext<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["ext_variance"]]))
ci_upper_ext<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["ext_variance"]]))
ci<- list(ci=data.frame(estimate=object$estimation[["estimate"]], ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level, adjust.method=adjust.method)
class(ci)<- c("confint.global", "twophase")
return(ci)
}# end of global-CI
} ## end of confint.twophase
# -----------------------------------------------------------------------------#
# -----------------------------------------------------------------------------#
#' @rdname confint
#' @export
confint.threephase<- function(object, parm, level = 0.95, adjust.method="none", ...){
# confint-method for for threephase estimations:
orig.level<- level
# --------------------------------#
# confint for threephase-smallarea:
if(is(object, "smallarea") & inherits(object, "threephase")){ # if class(object) is c("smallarea", "threephase")
#-----------#
# adjust confidence level if required:
if(adjust.method %in% c("none", "bonferroni")){
# bonferroni-correction:
if (adjust.method == "bonferroni"){
if(nrow(object$estimation)==1){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
} else {
level<- 1 - ((1 - level) / nrow(object$estimation))
}
}
} else {
stop("invalid argument for 'adjust.method'")
}
#-----------#
# calulating the quantile of the t-distribution:
# if synthetic-estimation was applied, the df's are calculated as n2 - # parameters of regression model:
if(object$input$method %in% c("synth", "psynth")){
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = unique(object$estimation[["n2"]]) - length(all.vars(object$input$formula.s1)[-1])))
# if non-synthetic-estimation was applied, the df's are calculated as n2G - 1:
} else {
t.val<- suppressWarnings(qt(p = 1-( (1-level)/2 ), df = object$estimation$n2G - 1))
}
#-----------#
# calculating the lower and upper confidence intervals based on the external variance:
ci_lower_g<- object$estimation$estimate - (t.val * sqrt(object$estimation$g_variance))
ci_upper_g<- object$estimation$estimate + (t.val * sqrt(object$estimation$g_variance))
# calculating the lower and upper confidence intervals based on the design-based variance:
ci_lower_ext<- object$estimation$estimate - (t.val * sqrt(object$estimation$ext_variance))
ci_upper_ext<- object$estimation$estimate + (t.val * sqrt(object$estimation$ext_variance))
ci<- list(ci=data.frame(area=object$estimation$area, estimate=object$estimation$estimate, ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level, adjust.method=adjust.method)
class(ci)<- c("confint.smallarea", "threephase")
return(ci)
} # end of smallarea-CI
# --------------------------------#
# summary for threephase-global:
if(is(object, "global") & inherits(object, "threephase")){ # if class(object) is c("global", "threephase")
if(adjust.method != "none"){
stop("'bonferroni-correction only reasonable for objects that contain a set of estimates'")
}
n2<- ifelse(object$input$cluster, object$samplesizes$n2_clust, object$samplesizes$n2)
t.val<- qt(p = 1-( (1-level)/2 ), df = n2 - length(all.vars(object$input$formula.s1)[-1])) # that's however not recommended by Daniel,
# see Article Crfr, Threephase, page 387
ci_lower_g<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["g_variance"]]))
ci_upper_g<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["g_variance"]]))
ci_lower_ext<- object$estimation[["estimate"]] - (t.val * sqrt(object$estimation[["ext_variance"]]))
ci_upper_ext<- object$estimation[["estimate"]] + (t.val * sqrt(object$estimation[["ext_variance"]]))
ci<- list(ci=data.frame(estimate=object$estimation[["estimate"]], ci_lower_ext, ci_upper_ext, ci_lower_g, ci_upper_g),
level=orig.level,adjust.method=adjust.method)
class(ci)<- c("confint.global", "threepase")
return(ci)
}# end of global-CI
}## end of confint.threephase
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.