R/coef.missingHE.R
In AnGabrio/missingHE: Missing Outcome Data in Health Economic Evaluation
Defines functions
coef.missingHE
Documented in
coef.missingHE
#' Extract regression coefficient estimates from objects in the class \code{missingHE}
#'
#' Produces a table printout with summary statistics for the regression coefficients of the health economic evaluation probabilistic model
#' run using the function \code{\link{selection}}, \code{\link{selection_long}}, \code{\link{pattern}} or \code{\link{hurdle}}.
#' @param object A \code{missingHE} object containing the results of the Bayesian modelling and the economic evaluation
#' @param prob A numeric vector of probabilities within the range (0,1), representing the upper and lower
#' CI sample quantiles to be calculated and returned for the estimates.
#' @param random Logical. If \code{random} is \code{TRUE}, the estimates of the random effects parameters are printed, when available.
#' @param time A number indicating the time point at which posterior results for the model coefficients should be reported (only for longitudinal models).
#' @param digits Number of digits to be displayed for each estimate.
#' @param ... Additional arguments affecting the summary produced.
#' @return Prints a table with some summary statistics, including posterior mean, standard deviation and lower and upper quantiles based on the
#' values specified in \code{prob}, for the posterior distributions of the regression coefficients of the effects and costs models run using the
#' function \code{selection}, \code{\link{selection_long}}, \code{pattern} or \code{hurdle}.
#' @seealso \code{\link{selection}} \code{\link{selection_long}} \code{\link{pattern}} \code{\link{hurdle}} \code{\link{diagnostic}} \code{\link{plot.missingHE}}
#' @author Andrea Gabrio
#' @importFrom stats quantile
#' @export
#' @examples
#' # For examples see the function \code{\link{selection}}, \code{\link{selection_long}},
#' # \code{\link{pattern}} or \code{\link{hurdle}}
#' #
#' #
coef.missingHE <- function(object, prob = c(0.025, 0.975), random = FALSE, time = 1, digits = 3, ...) {
exArgs <- list(...)
if(!inherits(object, what = "missingHE")) {
stop("Only objects of class 'missingHE' can be used")
}
if(length(prob) != 2 | is.numeric(prob) == FALSE | any(prob < 0) != FALSE | any(prob > 1) != FALSE) {
stop("You must provide valid lower/upper quantiles for the coefficients distributions")
}
if(object$data_format == "wide") {
if(length(grep("^SELECTION", object$model_output$type)) == 1 | length(grep("^HURDLE", object$model_output$type)) == 1) {
cov_e_fixed <- names(object$data_set$covariates_effects_fixed$Control)
p_e_fixed <- length(cov_e_fixed)
if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 2) {
mean_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , mean)
sd_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , sd)
ql_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , quantile, prob = prob[1])
qu_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 3) {
mean_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , mean)
sd_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , sd)
ql_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , quantile, prob = prob[1])
qu_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , quantile, prob = prob[2])
}
cov_c_fixed <- names(object$data_set$covariates_costs_fixed$Control)
p_c_fixed <- length(cov_c_fixed)
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 2 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 3 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 2 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 3 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_c_fixed <- "e"
mean_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2, mean)
sd_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2 , sd)
ql_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2 , quantile, prob = prob[1])
qu_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2 , quantile, prob = prob[2])
} else {dep_c_fixed <- NULL }
cov_e_random <- names(object$data_set$covariates_effects_random$Control)
p_e_random <- length(cov_e_random)
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 2) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[2])
}
cov_c_random <- names(object$data_set$covariates_costs_random$Control)
p_c_random <- length(cov_c_random)
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 2) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random)) == 3) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 0 & is.null(object$model_output$covariate_parameter_costs_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 2) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 3) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_c_random <- "e"
mean_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2, mean)
sd_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , sd)
ql_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[1])
qu_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[2])
mean_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2, mean)
sd_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , sd)
ql_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[1])
qu_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[2])
} else {dep_c_random <- NULL }
if(object$model_output$ind_fixed == FALSE) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
}
if(is.matrix(mean_cov_e_fixed) == FALSE) {
mean_cov_e_fixed_arm1 <- mean_cov_e_fixed[1]
sd_cov_e_fixed_arm1 <- sd_cov_e_fixed[1]
ql_cov_e_fixed_arm1 <- ql_cov_e_fixed[1]
qu_cov_e_fixed_arm1 <- qu_cov_e_fixed[1]
mean_cov_e_fixed_arm2 <- mean_cov_e_fixed[2]
sd_cov_e_fixed_arm2 <- sd_cov_e_fixed[2]
ql_cov_e_fixed_arm2 <- ql_cov_e_fixed[2]
qu_cov_e_fixed_arm2 <- qu_cov_e_fixed[2]
} else if(is.matrix(mean_cov_e_fixed) == TRUE) {
mean_cov_e_fixed_arm1 <- mean_cov_e_fixed[, 1]
sd_cov_e_fixed_arm1 <- sd_cov_e_fixed[, 1]
ql_cov_e_fixed_arm1 <- ql_cov_e_fixed[, 1]
qu_cov_e_fixed_arm1 <- qu_cov_e_fixed[, 1]
mean_cov_e_fixed_arm2 <- mean_cov_e_fixed[, 2]
sd_cov_e_fixed_arm2 <- sd_cov_e_fixed[, 2]
ql_cov_e_fixed_arm2 <- ql_cov_e_fixed[, 2]
qu_cov_e_fixed_arm2 <- qu_cov_e_fixed[, 2]
}
if(is.matrix(mean_cov_c_fixed) == FALSE) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[2]
} else if(is.matrix(mean_cov_c_fixed) == TRUE) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[, 1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[, 1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[, 1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[, 1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[, 2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[, 2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[, 2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[, 2]
}
}
if(length(grep("^PATTERN", object$model_output$type)) == 1) {
cov_e_fixed <- names(object$data_set$covariates_effects_fixed$Control)
p_e_fixed <- length(cov_e_fixed)
if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$control)) == 2) {
mean_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , mean)
sd_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , sd)
ql_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , quantile, prob = prob[1])
qu_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$control)) == 3) {
mean_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , mean)
sd_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , sd)
ql_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , quantile, prob = prob[1])
qu_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$intervention)) == 2) {
mean_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , mean)
sd_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , sd)
ql_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , quantile, prob = prob[1])
qu_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$intervention)) == 3) {
mean_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , mean)
sd_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , sd)
ql_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[1])
qu_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[2])
}
cov_c_fixed <- names(object$data_set$covariates_costs_fixed$Control)
p_c_fixed <- length(cov_c_fixed)
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control)) == 2) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control)) == 3) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention)) == 2) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention)) == 3) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1)) == 2) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1)) == 3) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , quantile, prob = prob[2])
}
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2)) == 2) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2)) == 3) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_c_fixed <- "e"
mean_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2, mean)
sd_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2 , sd)
ql_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2 , quantile, prob = prob[1])
qu_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2 , quantile, prob = prob[2])
mean_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2, mean)
sd_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2 , sd)
ql_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2 , quantile, prob = prob[1])
qu_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2 , quantile, prob = prob[2])
} else {dep_c_fixed <- NULL }
cov_e_random <- names(object$data_set$covariates_effects_random$Control)
p_e_random <- length(cov_e_random)
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 2) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[2])
}
cov_c_random <- names(object$data_set$covariates_costs_random$Control)
p_c_random <- length(cov_c_random)
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 2) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random)) == 3) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 0 & is.null(object$model_output$covariate_parameter_costs_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 2) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 3) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_c_random <- "e"
mean_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2, mean)
sd_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , sd)
ql_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[1])
qu_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[2])
mean_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2, mean)
sd_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , sd)
ql_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[1])
qu_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[2])
} else {dep_c_random <- NULL }
if(object$model_output$ind_fixed == FALSE) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c1_fixed <- rbind(mean_cov_c1_fixed, mean_dep_c1_fixed)
sd_cov_c1_fixed <- rbind(sd_cov_c1_fixed, sd_dep_c1_fixed)
ql_cov_c1_fixed <- rbind(ql_cov_c1_fixed, ql_dep_c1_fixed)
qu_cov_c1_fixed <- rbind(qu_cov_c1_fixed, qu_dep_c1_fixed)
mean_cov_c2_fixed <- rbind(mean_cov_c2_fixed, mean_dep_c2_fixed)
sd_cov_c2_fixed <- rbind(sd_cov_c2_fixed, sd_dep_c2_fixed)
ql_cov_c2_fixed <- rbind(ql_cov_c2_fixed, ql_dep_c2_fixed)
qu_cov_c2_fixed <- rbind(qu_cov_c2_fixed, qu_dep_c2_fixed)
}
mean_cov_e_fixed_arm1 <- mean_cov_e1_fixed
sd_cov_e_fixed_arm1 <- sd_cov_e1_fixed
ql_cov_e_fixed_arm1 <- ql_cov_e1_fixed
qu_cov_e_fixed_arm1 <- qu_cov_e1_fixed
mean_cov_e_fixed_arm2 <- mean_cov_e2_fixed
sd_cov_e_fixed_arm2 <- sd_cov_e2_fixed
ql_cov_e_fixed_arm2 <- ql_cov_e2_fixed
qu_cov_e_fixed_arm2 <- qu_cov_e2_fixed
mean_cov_c_fixed_arm1 <- mean_cov_c1_fixed
sd_cov_c_fixed_arm1 <- sd_cov_c1_fixed
ql_cov_c_fixed_arm1 <- ql_cov_c1_fixed
qu_cov_c_fixed_arm1 <- qu_cov_c1_fixed
mean_cov_c_fixed_arm2 <- mean_cov_c2_fixed
sd_cov_c_fixed_arm2 <- sd_cov_c2_fixed
ql_cov_c_fixed_arm2 <- ql_cov_c2_fixed
qu_cov_c_fixed_arm2 <- qu_cov_c2_fixed
}
if(length(grep("^PATTERN", object$model_output$type)) == 1) {
pat_arm1 <- length(unique(as.numeric(object$data_set$`patterns in comparator arm`)))
pat_arm2 <- length(unique(as.numeric(object$data_set$`patterns in reference arm`)))
pat_e1_index <- sort(rep(seq(1:pat_arm1), p_e_fixed))
pat_e2_index <- sort(rep(seq(1:pat_arm2), p_e_fixed))
table_e1_fixed <- matrix(NA, nrow = length(pat_e1_index), ncol = 4)
rownames(table_e1_fixed) <- paste(rep(cov_e_fixed, pat_arm1), pat_e1_index, sep = " pattern")
colnames(table_e1_fixed) <- c("mean", "sd", "lower", "upper")
table_e1_fixed[, 1] <- c(mean_cov_e_fixed_arm1)
table_e1_fixed[, 2] <- c(sd_cov_e_fixed_arm1)
table_e1_fixed[, 3] <- c(ql_cov_e_fixed_arm1)
table_e1_fixed[, 4] <- c(qu_cov_e_fixed_arm1)
table_e1_fixed <- round(table_e1_fixed, digits = digits)
table_e2_fixed <- matrix(NA, nrow = length(pat_e2_index), ncol = 4)
rownames(table_e2_fixed) <- paste(rep(cov_e_fixed, pat_arm2), pat_e2_index, sep = " pattern")
colnames(table_e2_fixed) <- c("mean", "sd", "lower", "upper")
table_e2_fixed[, 1] <- c(mean_cov_e_fixed_arm2)
table_e2_fixed[, 2] <- c(sd_cov_e_fixed_arm2)
table_e2_fixed[, 3] <- c(ql_cov_e_fixed_arm2)
table_e2_fixed[, 4] <- c(qu_cov_e_fixed_arm2)
table_e2_fixed <- round(table_e2_fixed, digits = digits)
pat_c1_index <- sort(rep(seq(1:pat_arm1), p_c_fixed))
pat_c2_index <- sort(rep(seq(1:pat_arm2), p_c_fixed))
table_c1_fixed <- matrix(NA, nrow = length(pat_c1_index), ncol = 4)
rownames(table_c1_fixed) <- paste(rep(cov_c_fixed, pat_arm1), pat_c1_index, sep = " pattern")
colnames(table_c1_fixed) <- c("mean", "sd", "lower", "upper")
table_c1_fixed[, 1] <- c(mean_cov_c_fixed_arm1)
table_c1_fixed[, 2] <- c(sd_cov_c_fixed_arm1)
table_c1_fixed[, 3] <- c(ql_cov_c_fixed_arm1)
table_c1_fixed[, 4] <- c(qu_cov_c_fixed_arm1)
table_c1_fixed <- round(table_c1_fixed, digits = digits)
table_c2_fixed <- matrix(NA, nrow = length(pat_c2_index), ncol = 4)
rownames(table_c2_fixed) <- paste(rep(cov_c_fixed, pat_arm2), pat_c2_index, sep = " pattern")
colnames(table_c2_fixed) <- c("mean", "sd", "lower", "upper")
table_c2_fixed[, 1] <- c(mean_cov_c_fixed_arm2)
table_c2_fixed[, 2] <- c(sd_cov_c_fixed_arm2)
table_c2_fixed[, 3] <- c(ql_cov_c_fixed_arm2)
table_c2_fixed[, 4] <- c(qu_cov_c_fixed_arm2)
table_c2_fixed <- round(table_c2_fixed, digits = digits)
}
if(length(grep("^SELECTION", object$model_output$type)) == 1 | length(grep("^HURDLE", object$model_output$type)) == 1) {
table_e1_fixed <- matrix(NA, nrow = p_e_fixed, ncol = 4)
rownames(table_e1_fixed) <- cov_e_fixed
colnames(table_e1_fixed) <- c("mean", "sd", "lower", "upper")
table_e1_fixed[, 1] <- c(mean_cov_e_fixed_arm1)
table_e1_fixed[, 2] <- c(sd_cov_e_fixed_arm1)
table_e1_fixed[, 3] <- c(ql_cov_e_fixed_arm1)
table_e1_fixed[, 4] <- c(qu_cov_e_fixed_arm1)
table_e1_fixed <- round(table_e1_fixed, digits = digits)
table_e2_fixed <- matrix(NA, nrow = p_e_fixed, ncol = 4)
rownames(table_e2_fixed) <- cov_e_fixed
colnames(table_e2_fixed) <- c("mean", "sd", "lower", "upper")
table_e2_fixed[, 1] <- c(mean_cov_e_fixed_arm2)
table_e2_fixed[, 2] <- c(sd_cov_e_fixed_arm2)
table_e2_fixed[, 3] <- c(ql_cov_e_fixed_arm2)
table_e2_fixed[, 4] <- c(qu_cov_e_fixed_arm2)
table_e2_fixed <- round(table_e2_fixed, digits = digits)
table_c1_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c1_fixed) <- cov_c_fixed
colnames(table_c1_fixed) <- c("mean", "sd", "lower", "upper")
table_c1_fixed[, 1] <- c(mean_cov_c_fixed_arm1)
table_c1_fixed[, 2] <- c(sd_cov_c_fixed_arm1)
table_c1_fixed[, 3] <- c(ql_cov_c_fixed_arm1)
table_c1_fixed[, 4] <- c(qu_cov_c_fixed_arm1)
table_c1_fixed <- round(table_c1_fixed, digits = digits)
table_c2_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c2_fixed) <- cov_c_fixed
colnames(table_c2_fixed) <- c("mean", "sd", "lower", "upper")
table_c2_fixed[, 1] <- c(mean_cov_c_fixed_arm2)
table_c2_fixed[, 2] <- c(sd_cov_c_fixed_arm2)
table_c2_fixed[, 3] <- c(ql_cov_c_fixed_arm2)
table_c2_fixed[, 4] <- c(qu_cov_c_fixed_arm2)
table_c2_fixed <- round(table_c2_fixed, digits = digits)
}
table_list_fixed <- list("Comparator" = list("Effects" = table_e1_fixed, "Costs" = table_c1_fixed),
"Reference" = list("Effects" = table_e2_fixed, "Costs" = table_c2_fixed))
table_list_random <- NULL
table_e1_random <- table_c1_random <- NULL
table_e2_random <- table_c2_random <- NULL
table_list_random <- list("Comparator" = list("Effects" = table_e1_random, "Costs" = table_c1_random),
"Reference" = list("Effects" = table_e2_random, "Costs" = table_c2_random))
clus_e_arm1 <- clus_e_arm2 <- NULL
clus_c_arm1 <- clus_c_arm2 <- NULL
if(p_c_random != 0 & object$model_output$ind_random == FALSE) {
cov_c_random <- c(cov_c_random, dep_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random)
} else if(p_c_random == 0 & object$model_output$ind_random == FALSE) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
pc_random <- 1
}
if(p_e_random != 0) {
mean_cov_e_random_arm1 <- mean_cov_e1_random
sd_cov_e_random_arm1 <- sd_cov_e1_random
ql_cov_e_random_arm1 <- ql_cov_e1_random
qu_cov_e_random_arm1 <- qu_cov_e1_random
mean_cov_e_random_arm2 <- mean_cov_e2_random
sd_cov_e_random_arm2 <- sd_cov_e2_random
ql_cov_e_random_arm2 <- ql_cov_e2_random
qu_cov_e_random_arm2 <- qu_cov_e2_random
clus_e_arm1 <- max(as.numeric(object$data_set$clus_effects$Control))
clus_e_arm2 <- max(as.numeric(object$data_set$clus_effects$Intervention))
clus_e1_index <- sort(rep(seq(1:clus_e_arm1), p_e_random))
clus_e2_index <- sort(rep(seq(1:clus_e_arm2), p_e_random))
table_e1_random <- matrix(NA, nrow = length(clus_e1_index), ncol = 4)
rownames(table_e1_random) <- paste(rep(cov_e_random, clus_e_arm1), clus_e1_index, sep = " ")
colnames(table_e1_random) <- c("mean", "sd", "lower", "upper")
table_e1_random[, 1] <- c(mean_cov_e_random_arm1)
table_e1_random[, 2] <- c(sd_cov_e_random_arm1)
table_e1_random[, 3] <- c(ql_cov_e_random_arm1)
table_e1_random[, 4] <- c(qu_cov_e_random_arm1)
table_e1_random <- round(table_e1_random, digits = digits)
table_e2_random <- matrix(NA, nrow = length(clus_e2_index), ncol = 4)
rownames(table_e2_random) <- paste(rep(cov_e_random, clus_e_arm2), clus_e2_index, sep = " ")
colnames(table_e2_random) <- c("mean", "sd", "lower", "upper")
table_e2_random[, 1] <- c(mean_cov_e_random_arm2)
table_e2_random[, 2] <- c(sd_cov_e_random_arm2)
table_e2_random[, 3] <- c(ql_cov_e_random_arm2)
table_e2_random[, 4] <- c(qu_cov_e_random_arm2)
table_e2_random <- round(table_e2_random, digits = digits)
table_list_random$Comparator$Effects <- table_e1_random
table_list_random$Reference$Effects <- table_e2_random
}
if(p_c_random != 0 | object$model_output$ind_random == FALSE) {
mean_cov_c_random_arm1 <- mean_cov_c1_random
sd_cov_c_random_arm1 <- sd_cov_c1_random
ql_cov_c_random_arm1 <- ql_cov_c1_random
qu_cov_c_random_arm1 <- qu_cov_c1_random
mean_cov_c_random_arm2 <- mean_cov_c2_random
sd_cov_c_random_arm2 <- sd_cov_c2_random
ql_cov_c_random_arm2 <- ql_cov_c2_random
qu_cov_c_random_arm2 <- qu_cov_c2_random
clus_c_arm1 <- max(as.numeric(object$data_set$clus_costs$Control))
clus_c_arm2 <- max(as.numeric(object$data_set$clus_costs$Intervention))
clus_c1_index <- sort(rep(seq(1:clus_c_arm1), p_c_random))
clus_c2_index <- sort(rep(seq(1:clus_c_arm2), p_c_random))
table_c1_random <- matrix(NA, nrow = length(clus_c1_index), ncol = 4)
rownames(table_c1_random) <- paste(rep(cov_c_random, clus_c_arm1), clus_c1_index, sep = " ")
colnames(table_c1_random) <- c("mean", "sd", "lower", "upper")
table_c1_random[, 1] <- c(mean_cov_c_random_arm1)
table_c1_random[, 2] <- c(sd_cov_c_random_arm1)
table_c1_random[, 3] <- c(ql_cov_c_random_arm1)
table_c1_random[, 4] <- c(qu_cov_c_random_arm1)
table_c1_random <- round(table_c1_random, digits = digits)
table_c2_random <- matrix(NA, nrow = length(clus_c2_index), ncol = 4)
rownames(table_c2_random) <- paste(rep(cov_c_random, clus_c_arm2), clus_c2_index, sep = " ")
colnames(table_c2_random) <- c("mean", "sd", "lower", "upper")
table_c2_random[, 1] <- c(mean_cov_c_random_arm2)
table_c2_random[, 2] <- c(sd_cov_c_random_arm2)
table_c2_random[, 3] <- c(ql_cov_c_random_arm2)
table_c2_random[, 4] <- c(qu_cov_c_random_arm2)
table_c2_random <- round(table_c2_random, digits = digits)
table_list_random$Comparator$Costs <- table_c1_random
table_list_random$Reference$Costs <- table_c2_random
}
if(random == TRUE & p_c_random == 0 & p_e_random == 0) {
stop("No random effects estimates found")
}
}
if(object$data_format == "long") {
max_time <- dim(object$data_set$effects$Control)[2]
time_dep <- object$time_dep
if(time < 1 | time > max_time | is.numeric(time) == FALSE | isTRUE(all.equal(time, as.integer(time))) == FALSE) {
stop("Time must be a numeric integer. Minimum time is 1 (baseline) and maximum time is the latest follow-up time in the data")
}
if(length(grep("^SELECTION", object$model_output$type)) == 1) {
cov_u_fixed <- names(object$data_set$covariates_effects_fixed$Control)
p_u_fixed <- length(cov_u_fixed)
if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 3) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 4) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 0) {
if(length(dim(object$model_output$covariate_parameter_effects_fixed$alpha)) == 3 & object$model_output$ind_fixed == FALSE) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed$alpha)) == 4 & object$model_output$ind_fixed == FALSE) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_dep_utime_u_fixed <- NULL
dep_dep_ctime_u_fixed <- NULL
if(object$model_output$ind_time_fixed == FALSE) {
if(time >= 2 & time_dep == "AR1") {
dep_dep_utime_u_fixed <- "u(j-1)"
mean_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, mean)
sd_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, sd)
ql_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_u_fixed <- "c(j-1)"
mean_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, mean)
sd_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, sd)
ql_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, quantile, prob = prob[2])
}
}
} else {
dep_dep_utime_u_fixed <- NULL
dep_dep_ctime_u_fixed <- NULL
}
cov_c_fixed <- names(object$data_set$covariates_costs_fixed$Control)
p_c_fixed <- length(cov_c_fixed)
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 3 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 4 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 3 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 4 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_c_fixed <- "u"
dep_dep_utime_c_fixed <- NULL
dep_dep_ctime_c_fixed <- NULL
mean_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, mean)
sd_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, sd)
ql_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, quantile, prob = prob[1])
qu_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, quantile, prob = prob[2])
if(object$model_output$ind_time_fixed == FALSE) {
if(time >= 2 & time_dep == "AR1") {
dep_dep_utime_c_fixed <- "u(j-1)"
mean_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, mean)
sd_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, sd)
ql_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_c_fixed <- "c(j-1)"
mean_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, mean)
sd_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, sd)
ql_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, quantile, prob = prob[2])
}
}
} else {
dep_c_fixed <- NULL
dep_dep_utime_c_fixed <- NULL
dep_dep_ctime_c_fixed <- NULL
}
cov_u_random <- names(object$data_set$covariates_effects_random$Control)
p_u_random <- length(cov_u_random)
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 4) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_effects_random)) == 0 & is.null(object$model_output$covariate_parameter_effects_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 4) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_dep_utime_u_random <- NULL
dep_dep_ctime_u_random <- NULL
if(is.null(object$model_output$covariate_parameter_effects_random$a1_tu) == FALSE) {
if(time >= 2 & time_dep == "AR1") {
dep_dep_utime_u_random <- "u(j-1)"
mean_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, mean)
sd_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, sd)
ql_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, quantile, prob = prob[2])
mean_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, mean)
sd_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, sd)
ql_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_u_random <- "c(j-1)"
mean_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, mean)
sd_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, sd)
ql_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, quantile, prob = prob[2])
mean_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, mean)
sd_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, sd)
ql_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, quantile, prob = prob[2])
}
} else if(is.null(object$model_output$covariate_parameter_costs_random$a1_tu) == TRUE) {
dep_dep_utime_u_random <- NULL
dep_dep_ctime_u_random <- NULL
}
} else {
dep_dep_utime_u_random <- NULL
dep_dep_ctime_u_random <- NULL
}
cov_c_random <- names(object$data_set$covariates_costs_random$Control)
p_c_random <- length(cov_c_random)
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 3) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random)) == 4) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 0 & is.null(object$model_output$covariate_parameter_costs_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 3) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 4) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_c_random <- "u"
mean_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, mean)
sd_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, sd)
ql_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, quantile, prob = prob[1])
qu_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, quantile, prob = prob[2])
mean_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, mean)
sd_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, sd)
ql_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, quantile, prob = prob[1])
qu_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, quantile, prob = prob[2])
dep_dep_utime_c_random <- NULL
dep_dep_ctime_c_random <- NULL
if(object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == FALSE) {
if(time >= 2) {
dep_dep_utime_c_random <- "u(j-1)"
mean_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, mean)
sd_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, sd)
ql_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, quantile, prob = prob[2])
mean_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, mean)
sd_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, sd)
ql_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_c_random <- "c(j-1)"
mean_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, mean)
sd_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, sd)
ql_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, quantile, prob = prob[2])
mean_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, mean)
sd_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, sd)
ql_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, quantile, prob = prob[2])
}
} else if(object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == TRUE) {
dep_dep_utime_c_random <- NULL
dep_dep_ctime_c_random <- NULL
}
} else {
dep_c_random <- NULL
dep_dep_utime_c_random <- NULL
dep_dep_ctime_c_random <- NULL
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_c_fixed != 0) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_c_fixed == 0) {
cov_c_fixed <- c(dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed)
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE & p_c_fixed != 0) {
if(time == 1) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
} else if(time > 1) {
if(time_dep == "AR1") {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed, dep_dep_utime_c_fixed, dep_dep_ctime_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed, mean_dep_utime_c_fixed, mean_dep_ctime_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed, sd_dep_utime_c_fixed, sd_dep_ctime_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed, ql_dep_utime_c_fixed, ql_dep_ctime_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed, qu_dep_utime_c_fixed, qu_dep_ctime_c_fixed)
} else {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
}
}
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE & p_c_fixed == 0) {
if(time == 1) {
cov_c_fixed <- c(dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed)
} else if(time > 1) {
if(time_dep == "AR1") {
cov_c_fixed <- c(dep_c_fixed, dep_dep_utime_c_fixed, dep_dep_ctime_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed, mean_dep_utime_c_fixed, mean_dep_ctime_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed, sd_dep_utime_c_fixed, sd_dep_ctime_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed, ql_dep_utime_c_fixed, ql_dep_ctime_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed, qu_dep_utime_c_fixed, qu_dep_ctime_c_fixed)
} else {
cov_c_fixed <- c(dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed)
}
}
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE) {
if(time == 1) {
cov_u_fixed <- cov_u_fixed
p_u_fixed <- length(cov_u_fixed)
mean_cov_u_fixed <- mean_cov_u_fixed
sd_cov_u_fixed <- sd_cov_u_fixed
ql_cov_u_fixed <- ql_cov_u_fixed
qu_cov_u_fixed <- qu_cov_u_fixed
} else if(time > 1) {
if(time_dep == "AR1") {
cov_u_fixed <- c(cov_u_fixed, dep_dep_utime_u_fixed, dep_dep_ctime_u_fixed)
p_u_fixed <- length(cov_u_fixed)
mean_cov_u_fixed <- rbind(mean_cov_u_fixed, mean_dep_utime_u_fixed, mean_dep_ctime_u_fixed)
sd_cov_u_fixed <- rbind(sd_cov_u_fixed, sd_dep_utime_u_fixed, sd_dep_ctime_u_fixed)
ql_cov_u_fixed <- rbind(ql_cov_u_fixed, ql_dep_utime_u_fixed, ql_dep_ctime_u_fixed)
qu_cov_u_fixed <- rbind(qu_cov_u_fixed, qu_dep_utime_u_fixed, qu_dep_ctime_u_fixed)
} else {
cov_u_fixed <- cov_u_fixed
p_u_fixed <- length(cov_u_fixed)
mean_cov_u_fixed <- mean_cov_u_fixed
sd_cov_u_fixed <- sd_cov_u_fixed
ql_cov_u_fixed <- ql_cov_u_fixed
qu_cov_u_fixed <- qu_cov_u_fixed
}
}
}
if(object$model_output$ind_fixed == TRUE & p_u_fixed == 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
} else if(object$model_output$ind_fixed == TRUE & p_u_fixed > 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[, 1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[, 1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[, 1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[, 1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[, 2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[, 2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[, 2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[, 2]
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_u_fixed == 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_u_fixed > 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[, 1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[, 1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[, 1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[, 1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[, 2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[, 2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[, 2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[, 2]
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE) {
if(time == 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
} else if(time > 1) {
if(time_dep == "AR1") {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[, 1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[, 1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[, 1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[, 1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[, 2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[, 2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[, 2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[, 2]
} else {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
}
}
}
if(object$model_output$ind_fixed == TRUE & p_c_fixed == 1) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[2]
} else if(object$model_output$ind_fixed == TRUE & p_c_fixed > 1) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[, 1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[, 1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[, 1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[, 1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[, 2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[, 2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[, 2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[, 2]
}
if(object$model_output$ind_fixed == FALSE & p_c_fixed != 0) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[, 1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[, 1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[, 1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[, 1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[, 2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[, 2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[, 2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[, 2]
} else if(object$model_output$ind_fixed == FALSE & p_c_fixed == 0) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[2]
}
}
if(length(grep("^SELECTION", object$model_output$type)) == 1) {
table_u1_fixed <- matrix(NA, nrow = p_u_fixed, ncol = 4)
rownames(table_u1_fixed) <- cov_u_fixed
colnames(table_u1_fixed) <- c("mean", "sd", "lower", "upper")
table_u1_fixed[, 1] <- c(mean_cov_u_fixed_arm1)
table_u1_fixed[, 2] <- c(sd_cov_u_fixed_arm1)
table_u1_fixed[, 3] <- c(ql_cov_u_fixed_arm1)
table_u1_fixed[, 4] <- c(qu_cov_u_fixed_arm1)
table_u1_fixed <- round(table_u1_fixed, digits = digits)
table_u2_fixed <- matrix(NA, nrow = p_u_fixed, ncol = 4)
rownames(table_u2_fixed) <- cov_u_fixed
colnames(table_u2_fixed) <- c("mean", "sd", "lower", "upper")
table_u2_fixed[, 1] <- c(mean_cov_u_fixed_arm2)
table_u2_fixed[, 2] <- c(sd_cov_u_fixed_arm2)
table_u2_fixed[, 3] <- c(ql_cov_u_fixed_arm2)
table_u2_fixed[, 4] <- c(qu_cov_u_fixed_arm2)
table_u2_fixed <- round(table_u2_fixed, digits = digits)
table_c1_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c1_fixed) <- cov_c_fixed
colnames(table_c1_fixed) <- c("mean", "sd", "lower", "upper")
table_c1_fixed[, 1] <- c(mean_cov_c_fixed_arm1)
table_c1_fixed[, 2] <- c(sd_cov_c_fixed_arm1)
table_c1_fixed[, 3] <- c(ql_cov_c_fixed_arm1)
table_c1_fixed[, 4] <- c(qu_cov_c_fixed_arm1)
table_c1_fixed <- round(table_c1_fixed, digits = digits)
table_c2_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c2_fixed) <- cov_c_fixed
colnames(table_c2_fixed) <- c("mean", "sd", "lower", "upper")
table_c2_fixed[, 1] <- c(mean_cov_c_fixed_arm2)
table_c2_fixed[, 2] <- c(sd_cov_c_fixed_arm2)
table_c2_fixed[, 3] <- c(ql_cov_c_fixed_arm2)
table_c2_fixed[, 4] <- c(qu_cov_c_fixed_arm2)
table_c2_fixed <- round(table_c2_fixed, digits = digits)
}
table_list_fixed <- list("Comparator" = list("Effects" = table_u1_fixed, "Costs" = table_c1_fixed),
"Reference" = list("Effects" = table_u2_fixed, "Costs" = table_c2_fixed))
table_list_random <- NULL
table_u1_random <- table_c1_random <- NULL
table_u2_random <- table_c2_random <- NULL
table_list_random <- list("Comparator" = list("Effects" = table_u1_random, "Costs" = table_c1_random),
"Reference" = list("Effects" = table_u2_random, "Costs" = table_c2_random))
clus_u_arm1 <- clus_u_arm2 <- NULL
clus_c_arm1 <- clus_c_arm2 <- NULL
if(p_c_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == TRUE) {
cov_c_random <- c(cov_c_random, dep_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random)
} else if(p_c_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == FALSE) {
if(time == 1) {
cov_c_random <- c(cov_c_random, dep_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random)
} else if(time > 1) {
cov_c_random <- c(cov_c_random, dep_c_random, dep_dep_utime_c_random, dep_dep_ctime_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random, mean_dep_utime_c1_random, mean_dep_ctime_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random, sd_dep_utime_c1_random, sd_dep_ctime_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random, ql_dep_utime_c1_random, ql_dep_ctime_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random, qu_dep_utime_c1_random, qu_dep_ctime_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random, mean_dep_utime_c2_random, mean_dep_ctime_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random, sd_dep_utime_c2_random, sd_dep_ctime_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random, ql_dep_utime_c2_random, ql_dep_ctime_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random, qu_dep_utime_c2_random, qu_dep_ctime_c2_random)
}
} else if(p_c_random == 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == TRUE) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
pc_random <- 1
} else if(p_c_random == 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == FALSE) {
if(is.null(dep_dep_utime_c_random) == FALSE) {
if(time == 1) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
} else if(time > 1) {
cov_c_random <- c(dep_c_random, dep_dep_utime_c_random, dep_dep_ctime_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_dep_c1_random, mean_dep_utime_c1_random, mean_dep_ctime_c1_random)
sd_cov_c1_random <- rbind(sd_dep_c1_random, sd_dep_utime_c1_random, sd_dep_ctime_c1_random)
ql_cov_c1_random <- rbind(ql_dep_c1_random, ql_dep_utime_c1_random, ql_dep_ctime_c1_random)
qu_cov_c1_random <- rbind(qu_dep_c1_random, qu_dep_utime_c1_random, qu_dep_ctime_c1_random)
mean_cov_c2_random <- rbind(mean_dep_c2_random, mean_dep_utime_c2_random, mean_dep_ctime_c2_random)
sd_cov_c2_random <- rbind(sd_dep_c2_random, sd_dep_utime_c2_random, sd_dep_ctime_c2_random)
ql_cov_c2_random <- rbind(ql_dep_c2_random, ql_dep_utime_c2_random, ql_dep_ctime_c2_random)
qu_cov_c2_random <- rbind(qu_dep_c2_random, qu_dep_utime_c2_random, qu_dep_ctime_c2_random)
}
} else if(is.null(dep_dep_utime_c_random) == TRUE) {
if(time == 1) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
} else if(time > 1) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
}
}
}
if(p_u_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$a1_tu) == FALSE) {
if(time > 1 & time_dep == "AR1") {
cov_u_random <- c(cov_u_random, dep_dep_utime_u_random, dep_dep_ctime_u_random)
p_u_random <- length(cov_u_random)
mean_cov_u1_random <- rbind(mean_cov_u1_random, mean_dep_utime_u1_random, mean_dep_ctime_u1_random)
sd_cov_u1_random <- rbind(sd_cov_u1_random, sd_dep_utime_u1_random, sd_dep_ctime_u1_random)
ql_cov_u1_random <- rbind(ql_cov_u1_random, ql_dep_utime_u1_random, ql_dep_ctime_u1_random)
qu_cov_u1_random <- rbind(qu_cov_u1_random, qu_dep_utime_u1_random, qu_dep_ctime_u1_random)
mean_cov_u2_random <- rbind(mean_cov_u2_random, mean_dep_utime_u2_random, mean_dep_ctime_u2_random)
sd_cov_u2_random <- rbind(sd_cov_u2_random, sd_dep_utime_u2_random, sd_dep_ctime_u2_random)
ql_cov_u2_random <- rbind(ql_cov_u2_random, ql_dep_utime_u2_random, ql_dep_ctime_u2_random)
qu_cov_u2_random <- rbind(qu_cov_u2_random, qu_dep_utime_u2_random, qu_dep_ctime_u2_random)
}
} else if (p_u_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$a1_tu) == TRUE) {
if(time > 1 & time_dep == "AR1") {
cov_u_random <- c(cov_u_random, dep_dep_utime_u_random, dep_dep_ctime_u_random)
p_u_random <- length(cov_u_random)
mean_cov_u1_random <- rbind(mean_cov_u1_random, mean_dep_utime_u1_random, mean_dep_ctime_u1_random)
sd_cov_u1_random <- rbind(sd_cov_u1_random, sd_dep_utime_u1_random, sd_dep_ctime_u1_random)
ql_cov_u1_random <- rbind(ql_cov_u1_random, ql_dep_utime_u1_random, ql_dep_ctime_u1_random)
qu_cov_u1_random <- rbind(qu_cov_u1_random, qu_dep_utime_u1_random, qu_dep_ctime_u1_random)
mean_cov_u2_random <- rbind(mean_cov_u2_random, mean_dep_utime_u2_random, mean_dep_ctime_u2_random)
sd_cov_u2_random <- rbind(sd_cov_u2_random, sd_dep_utime_u2_random, sd_dep_ctime_u2_random)
ql_cov_u2_random <- rbind(ql_cov_u2_random, ql_dep_utime_u2_random, ql_dep_ctime_u2_random)
qu_cov_u2_random <- rbind(qu_cov_u2_random, qu_dep_utime_u2_random, qu_dep_ctime_u2_random)
}
}
if(p_u_random != 0) {
mean_cov_u_random_arm1 <- mean_cov_u1_random
sd_cov_u_random_arm1 <- sd_cov_u1_random
ql_cov_u_random_arm1 <- ql_cov_u1_random
qu_cov_u_random_arm1 <- qu_cov_u1_random
mean_cov_u_random_arm2 <- mean_cov_u2_random
sd_cov_u_random_arm2 <- sd_cov_u2_random
ql_cov_u_random_arm2 <- ql_cov_u2_random
qu_cov_u_random_arm2 <- qu_cov_u2_random
clus_u_arm1 <- max(as.numeric(object$data_set$clus_effects$Control))
clus_u_arm2 <- max(as.numeric(object$data_set$clus_effects$Intervention))
clus_u1_index <- sort(rep(seq(1:clus_u_arm1), p_u_random))
clus_u2_index <- sort(rep(seq(1:clus_u_arm2), p_u_random))
table_u1_random <- matrix(NA, nrow = length(clus_u1_index), ncol = 4)
rownames(table_u1_random) <- paste(rep(cov_u_random, clus_u_arm1), clus_u1_index, sep = " ")
colnames(table_u1_random) <- c("mean", "sd", "lower", "upper")
table_u1_random[, 1] <- c(mean_cov_u_random_arm1)
table_u1_random[, 2] <- c(sd_cov_u_random_arm1)
table_u1_random[, 3] <- c(ql_cov_u_random_arm1)
table_u1_random[, 4] <- c(qu_cov_u_random_arm1)
table_u1_random <- round(table_u1_random, digits = digits)
table_u2_random <- matrix(NA, nrow = length(clus_u2_index), ncol = 4)
rownames(table_u2_random) <- paste(rep(cov_u_random, clus_u_arm2), clus_u2_index, sep = " ")
colnames(table_u2_random) <- c("mean", "sd", "lower", "upper")
table_u2_random[, 1] <- c(mean_cov_u_random_arm2)
table_u2_random[, 2] <- c(sd_cov_u_random_arm2)
table_u2_random[, 3] <- c(ql_cov_u_random_arm2)
table_u2_random[, 4] <- c(qu_cov_u_random_arm2)
table_u2_random <- round(table_u2_random, digits = digits)
table_list_random$Comparator$Effects <- table_u1_random
table_list_random$Reference$Effects <- table_u2_random
}
if(p_c_random != 0 | object$model_output$ind_random == FALSE) {
mean_cov_c_random_arm1 <- mean_cov_c1_random
sd_cov_c_random_arm1 <- sd_cov_c1_random
ql_cov_c_random_arm1 <- ql_cov_c1_random
qu_cov_c_random_arm1 <- qu_cov_c1_random
mean_cov_c_random_arm2 <- mean_cov_c2_random
sd_cov_c_random_arm2 <- sd_cov_c2_random
ql_cov_c_random_arm2 <- ql_cov_c2_random
qu_cov_c_random_arm2 <- qu_cov_c2_random
clus_c_arm1 <- max(as.numeric(object$data_set$clus_costs$Control))
clus_c_arm2 <- max(as.numeric(object$data_set$clus_costs$Intervention))
clus_c1_index <- sort(rep(seq(1:clus_c_arm1), p_c_random))
clus_c2_index <- sort(rep(seq(1:clus_c_arm2), p_c_random))
table_c1_random <- matrix(NA, nrow = length(clus_c1_index), ncol = 4)
rownames(table_c1_random) <- paste(rep(cov_c_random, clus_c_arm1), clus_c1_index, sep = " ")
colnames(table_c1_random) <- c("mean", "sd", "lower", "upper")
table_c1_random[, 1] <- c(mean_cov_c_random_arm1)
table_c1_random[, 2] <- c(sd_cov_c_random_arm1)
table_c1_random[, 3] <- c(ql_cov_c_random_arm1)
table_c1_random[, 4] <- c(qu_cov_c_random_arm1)
table_c1_random <- round(table_c1_random, digits = digits)
table_c2_random <- matrix(NA, nrow = length(clus_c2_index), ncol = 4)
rownames(table_c2_random) <- paste(rep(cov_c_random, clus_c_arm2), clus_c2_index, sep = " ")
colnames(table_c2_random) <- c("mean", "sd", "lower", "upper")
table_c2_random[, 1] <- c(mean_cov_c_random_arm2)
table_c2_random[, 2] <- c(sd_cov_c_random_arm2)
table_c2_random[, 3] <- c(ql_cov_c_random_arm2)
table_c2_random[, 4] <- c(qu_cov_c_random_arm2)
table_c2_random <- round(table_c2_random, digits = digits)
table_list_random$Comparator$Costs <- table_c1_random
table_list_random$Reference$Costs <- table_c2_random
}
if(random == TRUE & p_c_random == 0 & p_u_random == 0) {
stop("No random effects estimates found for effect and cost models")
}
}
if(random == FALSE) {
print(table_list_fixed)
} else if(random == TRUE) {
print(table_list_random)
}
}
AnGabrio/missingHE documentation built on March 22, 2023, 12:55 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 coef.missingHE
Documented in coef.missingHE
#' Extract regression coefficient estimates from objects in the class \code{missingHE}
#'
#' Produces a table printout with summary statistics for the regression coefficients of the health economic evaluation probabilistic model
#' run using the function \code{\link{selection}}, \code{\link{selection_long}}, \code{\link{pattern}} or \code{\link{hurdle}}.
#' @param object A \code{missingHE} object containing the results of the Bayesian modelling and the economic evaluation
#' @param prob A numeric vector of probabilities within the range (0,1), representing the upper and lower
#' CI sample quantiles to be calculated and returned for the estimates.
#' @param random Logical. If \code{random} is \code{TRUE}, the estimates of the random effects parameters are printed, when available.
#' @param time A number indicating the time point at which posterior results for the model coefficients should be reported (only for longitudinal models).
#' @param digits Number of digits to be displayed for each estimate.
#' @param ... Additional arguments affecting the summary produced.
#' @return Prints a table with some summary statistics, including posterior mean, standard deviation and lower and upper quantiles based on the
#' values specified in \code{prob}, for the posterior distributions of the regression coefficients of the effects and costs models run using the
#' function \code{selection}, \code{\link{selection_long}}, \code{pattern} or \code{hurdle}.
#' @seealso \code{\link{selection}} \code{\link{selection_long}} \code{\link{pattern}} \code{\link{hurdle}} \code{\link{diagnostic}} \code{\link{plot.missingHE}}
#' @author Andrea Gabrio
#' @importFrom stats quantile
#' @export
#' @examples
#' # For examples see the function \code{\link{selection}}, \code{\link{selection_long}},
#' # \code{\link{pattern}} or \code{\link{hurdle}}
#' #
#' #
coef.missingHE <- function(object, prob = c(0.025, 0.975), random = FALSE, time = 1, digits = 3, ...) {
exArgs <- list(...)
if(!inherits(object, what = "missingHE")) {
stop("Only objects of class 'missingHE' can be used")
}
if(length(prob) != 2 | is.numeric(prob) == FALSE | any(prob < 0) != FALSE | any(prob > 1) != FALSE) {
stop("You must provide valid lower/upper quantiles for the coefficients distributions")
}
if(object$data_format == "wide") {
if(length(grep("^SELECTION", object$model_output$type)) == 1 | length(grep("^HURDLE", object$model_output$type)) == 1) {
cov_e_fixed <- names(object$data_set$covariates_effects_fixed$Control)
p_e_fixed <- length(cov_e_fixed)
if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 2) {
mean_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , mean)
sd_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , sd)
ql_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , quantile, prob = prob[1])
qu_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 3) {
mean_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , mean)
sd_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , sd)
ql_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , quantile, prob = prob[1])
qu_cov_e_fixed <- apply(object$model_output$covariate_parameter_effects_fixed, c(2, 3) , quantile, prob = prob[2])
}
cov_c_fixed <- names(object$data_set$covariates_costs_fixed$Control)
p_c_fixed <- length(cov_c_fixed)
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 2 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 3 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 2 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 3 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_c_fixed <- "e"
mean_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2, mean)
sd_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2 , sd)
ql_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2 , quantile, prob = prob[1])
qu_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f, 2 , quantile, prob = prob[2])
} else {dep_c_fixed <- NULL }
cov_e_random <- names(object$data_set$covariates_effects_random$Control)
p_e_random <- length(cov_e_random)
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 2) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[2])
}
cov_c_random <- names(object$data_set$covariates_costs_random$Control)
p_c_random <- length(cov_c_random)
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 2) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random)) == 3) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 0 & is.null(object$model_output$covariate_parameter_costs_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 2) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 3) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_c_random <- "e"
mean_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2, mean)
sd_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , sd)
ql_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[1])
qu_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[2])
mean_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2, mean)
sd_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , sd)
ql_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[1])
qu_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[2])
} else {dep_c_random <- NULL }
if(object$model_output$ind_fixed == FALSE) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
}
if(is.matrix(mean_cov_e_fixed) == FALSE) {
mean_cov_e_fixed_arm1 <- mean_cov_e_fixed[1]
sd_cov_e_fixed_arm1 <- sd_cov_e_fixed[1]
ql_cov_e_fixed_arm1 <- ql_cov_e_fixed[1]
qu_cov_e_fixed_arm1 <- qu_cov_e_fixed[1]
mean_cov_e_fixed_arm2 <- mean_cov_e_fixed[2]
sd_cov_e_fixed_arm2 <- sd_cov_e_fixed[2]
ql_cov_e_fixed_arm2 <- ql_cov_e_fixed[2]
qu_cov_e_fixed_arm2 <- qu_cov_e_fixed[2]
} else if(is.matrix(mean_cov_e_fixed) == TRUE) {
mean_cov_e_fixed_arm1 <- mean_cov_e_fixed[, 1]
sd_cov_e_fixed_arm1 <- sd_cov_e_fixed[, 1]
ql_cov_e_fixed_arm1 <- ql_cov_e_fixed[, 1]
qu_cov_e_fixed_arm1 <- qu_cov_e_fixed[, 1]
mean_cov_e_fixed_arm2 <- mean_cov_e_fixed[, 2]
sd_cov_e_fixed_arm2 <- sd_cov_e_fixed[, 2]
ql_cov_e_fixed_arm2 <- ql_cov_e_fixed[, 2]
qu_cov_e_fixed_arm2 <- qu_cov_e_fixed[, 2]
}
if(is.matrix(mean_cov_c_fixed) == FALSE) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[2]
} else if(is.matrix(mean_cov_c_fixed) == TRUE) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[, 1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[, 1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[, 1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[, 1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[, 2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[, 2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[, 2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[, 2]
}
}
if(length(grep("^PATTERN", object$model_output$type)) == 1) {
cov_e_fixed <- names(object$data_set$covariates_effects_fixed$Control)
p_e_fixed <- length(cov_e_fixed)
if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$control)) == 2) {
mean_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , mean)
sd_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , sd)
ql_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , quantile, prob = prob[1])
qu_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$control)) == 3) {
mean_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , mean)
sd_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , sd)
ql_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , quantile, prob = prob[1])
qu_cov_e1_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$control, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$intervention)) == 2) {
mean_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , mean)
sd_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , sd)
ql_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , quantile, prob = prob[1])
qu_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed_pattern$intervention)) == 3) {
mean_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , mean)
sd_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , sd)
ql_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[1])
qu_cov_e2_fixed <- apply(object$model_output$covariate_parameter_effects_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[2])
}
cov_c_fixed <- names(object$data_set$covariates_costs_fixed$Control)
p_c_fixed <- length(cov_c_fixed)
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control)) == 2) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control)) == 3) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention)) == 2) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention)) == 3) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1)) == 2) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1)) == 3) {
mean_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , mean)
sd_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , sd)
ql_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_p1, c(2, 3) , quantile, prob = prob[2])
}
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2)) == 2) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2)) == 3) {
mean_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , mean)
sd_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , sd)
ql_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_p2, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_c_fixed <- "e"
mean_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2, mean)
sd_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2 , sd)
ql_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2 , quantile, prob = prob[1])
qu_dep_c1_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$control$beta_f_p1, 2 , quantile, prob = prob[2])
mean_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2, mean)
sd_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2 , sd)
ql_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2 , quantile, prob = prob[1])
qu_dep_c2_fixed <- apply(object$model_output$covariate_parameter_costs_fixed_pattern$intervention$beta_f_p2, 2 , quantile, prob = prob[2])
} else {dep_c_fixed <- NULL }
cov_e_random <- names(object$data_set$covariates_effects_random$Control)
p_e_random <- length(cov_e_random)
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 2) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, 2 , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , mean)
sd_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , sd)
ql_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[1])
qu_cov_e1_random <- apply(object$model_output$covariate_parameter_effects_random$a1, c(2, 3) , quantile, prob = prob[2])
mean_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , mean)
sd_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , sd)
ql_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[1])
qu_cov_e2_random <- apply(object$model_output$covariate_parameter_effects_random$a2, c(2, 3) , quantile, prob = prob[2])
}
cov_c_random <- names(object$data_set$covariates_costs_random$Control)
p_c_random <- length(cov_c_random)
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 2) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random)) == 3) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random, c(2, 3) , quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 0 & is.null(object$model_output$covariate_parameter_costs_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 2) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, 2 , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, 2 , quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 3) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1, c(2, 3) , quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2, c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_c_random <- "e"
mean_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2, mean)
sd_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , sd)
ql_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[1])
qu_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f, 2 , quantile, prob = prob[2])
mean_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2, mean)
sd_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , sd)
ql_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[1])
qu_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f, 2 , quantile, prob = prob[2])
} else {dep_c_random <- NULL }
if(object$model_output$ind_fixed == FALSE) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c1_fixed <- rbind(mean_cov_c1_fixed, mean_dep_c1_fixed)
sd_cov_c1_fixed <- rbind(sd_cov_c1_fixed, sd_dep_c1_fixed)
ql_cov_c1_fixed <- rbind(ql_cov_c1_fixed, ql_dep_c1_fixed)
qu_cov_c1_fixed <- rbind(qu_cov_c1_fixed, qu_dep_c1_fixed)
mean_cov_c2_fixed <- rbind(mean_cov_c2_fixed, mean_dep_c2_fixed)
sd_cov_c2_fixed <- rbind(sd_cov_c2_fixed, sd_dep_c2_fixed)
ql_cov_c2_fixed <- rbind(ql_cov_c2_fixed, ql_dep_c2_fixed)
qu_cov_c2_fixed <- rbind(qu_cov_c2_fixed, qu_dep_c2_fixed)
}
mean_cov_e_fixed_arm1 <- mean_cov_e1_fixed
sd_cov_e_fixed_arm1 <- sd_cov_e1_fixed
ql_cov_e_fixed_arm1 <- ql_cov_e1_fixed
qu_cov_e_fixed_arm1 <- qu_cov_e1_fixed
mean_cov_e_fixed_arm2 <- mean_cov_e2_fixed
sd_cov_e_fixed_arm2 <- sd_cov_e2_fixed
ql_cov_e_fixed_arm2 <- ql_cov_e2_fixed
qu_cov_e_fixed_arm2 <- qu_cov_e2_fixed
mean_cov_c_fixed_arm1 <- mean_cov_c1_fixed
sd_cov_c_fixed_arm1 <- sd_cov_c1_fixed
ql_cov_c_fixed_arm1 <- ql_cov_c1_fixed
qu_cov_c_fixed_arm1 <- qu_cov_c1_fixed
mean_cov_c_fixed_arm2 <- mean_cov_c2_fixed
sd_cov_c_fixed_arm2 <- sd_cov_c2_fixed
ql_cov_c_fixed_arm2 <- ql_cov_c2_fixed
qu_cov_c_fixed_arm2 <- qu_cov_c2_fixed
}
if(length(grep("^PATTERN", object$model_output$type)) == 1) {
pat_arm1 <- length(unique(as.numeric(object$data_set$`patterns in comparator arm`)))
pat_arm2 <- length(unique(as.numeric(object$data_set$`patterns in reference arm`)))
pat_e1_index <- sort(rep(seq(1:pat_arm1), p_e_fixed))
pat_e2_index <- sort(rep(seq(1:pat_arm2), p_e_fixed))
table_e1_fixed <- matrix(NA, nrow = length(pat_e1_index), ncol = 4)
rownames(table_e1_fixed) <- paste(rep(cov_e_fixed, pat_arm1), pat_e1_index, sep = " pattern")
colnames(table_e1_fixed) <- c("mean", "sd", "lower", "upper")
table_e1_fixed[, 1] <- c(mean_cov_e_fixed_arm1)
table_e1_fixed[, 2] <- c(sd_cov_e_fixed_arm1)
table_e1_fixed[, 3] <- c(ql_cov_e_fixed_arm1)
table_e1_fixed[, 4] <- c(qu_cov_e_fixed_arm1)
table_e1_fixed <- round(table_e1_fixed, digits = digits)
table_e2_fixed <- matrix(NA, nrow = length(pat_e2_index), ncol = 4)
rownames(table_e2_fixed) <- paste(rep(cov_e_fixed, pat_arm2), pat_e2_index, sep = " pattern")
colnames(table_e2_fixed) <- c("mean", "sd", "lower", "upper")
table_e2_fixed[, 1] <- c(mean_cov_e_fixed_arm2)
table_e2_fixed[, 2] <- c(sd_cov_e_fixed_arm2)
table_e2_fixed[, 3] <- c(ql_cov_e_fixed_arm2)
table_e2_fixed[, 4] <- c(qu_cov_e_fixed_arm2)
table_e2_fixed <- round(table_e2_fixed, digits = digits)
pat_c1_index <- sort(rep(seq(1:pat_arm1), p_c_fixed))
pat_c2_index <- sort(rep(seq(1:pat_arm2), p_c_fixed))
table_c1_fixed <- matrix(NA, nrow = length(pat_c1_index), ncol = 4)
rownames(table_c1_fixed) <- paste(rep(cov_c_fixed, pat_arm1), pat_c1_index, sep = " pattern")
colnames(table_c1_fixed) <- c("mean", "sd", "lower", "upper")
table_c1_fixed[, 1] <- c(mean_cov_c_fixed_arm1)
table_c1_fixed[, 2] <- c(sd_cov_c_fixed_arm1)
table_c1_fixed[, 3] <- c(ql_cov_c_fixed_arm1)
table_c1_fixed[, 4] <- c(qu_cov_c_fixed_arm1)
table_c1_fixed <- round(table_c1_fixed, digits = digits)
table_c2_fixed <- matrix(NA, nrow = length(pat_c2_index), ncol = 4)
rownames(table_c2_fixed) <- paste(rep(cov_c_fixed, pat_arm2), pat_c2_index, sep = " pattern")
colnames(table_c2_fixed) <- c("mean", "sd", "lower", "upper")
table_c2_fixed[, 1] <- c(mean_cov_c_fixed_arm2)
table_c2_fixed[, 2] <- c(sd_cov_c_fixed_arm2)
table_c2_fixed[, 3] <- c(ql_cov_c_fixed_arm2)
table_c2_fixed[, 4] <- c(qu_cov_c_fixed_arm2)
table_c2_fixed <- round(table_c2_fixed, digits = digits)
}
if(length(grep("^SELECTION", object$model_output$type)) == 1 | length(grep("^HURDLE", object$model_output$type)) == 1) {
table_e1_fixed <- matrix(NA, nrow = p_e_fixed, ncol = 4)
rownames(table_e1_fixed) <- cov_e_fixed
colnames(table_e1_fixed) <- c("mean", "sd", "lower", "upper")
table_e1_fixed[, 1] <- c(mean_cov_e_fixed_arm1)
table_e1_fixed[, 2] <- c(sd_cov_e_fixed_arm1)
table_e1_fixed[, 3] <- c(ql_cov_e_fixed_arm1)
table_e1_fixed[, 4] <- c(qu_cov_e_fixed_arm1)
table_e1_fixed <- round(table_e1_fixed, digits = digits)
table_e2_fixed <- matrix(NA, nrow = p_e_fixed, ncol = 4)
rownames(table_e2_fixed) <- cov_e_fixed
colnames(table_e2_fixed) <- c("mean", "sd", "lower", "upper")
table_e2_fixed[, 1] <- c(mean_cov_e_fixed_arm2)
table_e2_fixed[, 2] <- c(sd_cov_e_fixed_arm2)
table_e2_fixed[, 3] <- c(ql_cov_e_fixed_arm2)
table_e2_fixed[, 4] <- c(qu_cov_e_fixed_arm2)
table_e2_fixed <- round(table_e2_fixed, digits = digits)
table_c1_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c1_fixed) <- cov_c_fixed
colnames(table_c1_fixed) <- c("mean", "sd", "lower", "upper")
table_c1_fixed[, 1] <- c(mean_cov_c_fixed_arm1)
table_c1_fixed[, 2] <- c(sd_cov_c_fixed_arm1)
table_c1_fixed[, 3] <- c(ql_cov_c_fixed_arm1)
table_c1_fixed[, 4] <- c(qu_cov_c_fixed_arm1)
table_c1_fixed <- round(table_c1_fixed, digits = digits)
table_c2_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c2_fixed) <- cov_c_fixed
colnames(table_c2_fixed) <- c("mean", "sd", "lower", "upper")
table_c2_fixed[, 1] <- c(mean_cov_c_fixed_arm2)
table_c2_fixed[, 2] <- c(sd_cov_c_fixed_arm2)
table_c2_fixed[, 3] <- c(ql_cov_c_fixed_arm2)
table_c2_fixed[, 4] <- c(qu_cov_c_fixed_arm2)
table_c2_fixed <- round(table_c2_fixed, digits = digits)
}
table_list_fixed <- list("Comparator" = list("Effects" = table_e1_fixed, "Costs" = table_c1_fixed),
"Reference" = list("Effects" = table_e2_fixed, "Costs" = table_c2_fixed))
table_list_random <- NULL
table_e1_random <- table_c1_random <- NULL
table_e2_random <- table_c2_random <- NULL
table_list_random <- list("Comparator" = list("Effects" = table_e1_random, "Costs" = table_c1_random),
"Reference" = list("Effects" = table_e2_random, "Costs" = table_c2_random))
clus_e_arm1 <- clus_e_arm2 <- NULL
clus_c_arm1 <- clus_c_arm2 <- NULL
if(p_c_random != 0 & object$model_output$ind_random == FALSE) {
cov_c_random <- c(cov_c_random, dep_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random)
} else if(p_c_random == 0 & object$model_output$ind_random == FALSE) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
pc_random <- 1
}
if(p_e_random != 0) {
mean_cov_e_random_arm1 <- mean_cov_e1_random
sd_cov_e_random_arm1 <- sd_cov_e1_random
ql_cov_e_random_arm1 <- ql_cov_e1_random
qu_cov_e_random_arm1 <- qu_cov_e1_random
mean_cov_e_random_arm2 <- mean_cov_e2_random
sd_cov_e_random_arm2 <- sd_cov_e2_random
ql_cov_e_random_arm2 <- ql_cov_e2_random
qu_cov_e_random_arm2 <- qu_cov_e2_random
clus_e_arm1 <- max(as.numeric(object$data_set$clus_effects$Control))
clus_e_arm2 <- max(as.numeric(object$data_set$clus_effects$Intervention))
clus_e1_index <- sort(rep(seq(1:clus_e_arm1), p_e_random))
clus_e2_index <- sort(rep(seq(1:clus_e_arm2), p_e_random))
table_e1_random <- matrix(NA, nrow = length(clus_e1_index), ncol = 4)
rownames(table_e1_random) <- paste(rep(cov_e_random, clus_e_arm1), clus_e1_index, sep = " ")
colnames(table_e1_random) <- c("mean", "sd", "lower", "upper")
table_e1_random[, 1] <- c(mean_cov_e_random_arm1)
table_e1_random[, 2] <- c(sd_cov_e_random_arm1)
table_e1_random[, 3] <- c(ql_cov_e_random_arm1)
table_e1_random[, 4] <- c(qu_cov_e_random_arm1)
table_e1_random <- round(table_e1_random, digits = digits)
table_e2_random <- matrix(NA, nrow = length(clus_e2_index), ncol = 4)
rownames(table_e2_random) <- paste(rep(cov_e_random, clus_e_arm2), clus_e2_index, sep = " ")
colnames(table_e2_random) <- c("mean", "sd", "lower", "upper")
table_e2_random[, 1] <- c(mean_cov_e_random_arm2)
table_e2_random[, 2] <- c(sd_cov_e_random_arm2)
table_e2_random[, 3] <- c(ql_cov_e_random_arm2)
table_e2_random[, 4] <- c(qu_cov_e_random_arm2)
table_e2_random <- round(table_e2_random, digits = digits)
table_list_random$Comparator$Effects <- table_e1_random
table_list_random$Reference$Effects <- table_e2_random
}
if(p_c_random != 0 | object$model_output$ind_random == FALSE) {
mean_cov_c_random_arm1 <- mean_cov_c1_random
sd_cov_c_random_arm1 <- sd_cov_c1_random
ql_cov_c_random_arm1 <- ql_cov_c1_random
qu_cov_c_random_arm1 <- qu_cov_c1_random
mean_cov_c_random_arm2 <- mean_cov_c2_random
sd_cov_c_random_arm2 <- sd_cov_c2_random
ql_cov_c_random_arm2 <- ql_cov_c2_random
qu_cov_c_random_arm2 <- qu_cov_c2_random
clus_c_arm1 <- max(as.numeric(object$data_set$clus_costs$Control))
clus_c_arm2 <- max(as.numeric(object$data_set$clus_costs$Intervention))
clus_c1_index <- sort(rep(seq(1:clus_c_arm1), p_c_random))
clus_c2_index <- sort(rep(seq(1:clus_c_arm2), p_c_random))
table_c1_random <- matrix(NA, nrow = length(clus_c1_index), ncol = 4)
rownames(table_c1_random) <- paste(rep(cov_c_random, clus_c_arm1), clus_c1_index, sep = " ")
colnames(table_c1_random) <- c("mean", "sd", "lower", "upper")
table_c1_random[, 1] <- c(mean_cov_c_random_arm1)
table_c1_random[, 2] <- c(sd_cov_c_random_arm1)
table_c1_random[, 3] <- c(ql_cov_c_random_arm1)
table_c1_random[, 4] <- c(qu_cov_c_random_arm1)
table_c1_random <- round(table_c1_random, digits = digits)
table_c2_random <- matrix(NA, nrow = length(clus_c2_index), ncol = 4)
rownames(table_c2_random) <- paste(rep(cov_c_random, clus_c_arm2), clus_c2_index, sep = " ")
colnames(table_c2_random) <- c("mean", "sd", "lower", "upper")
table_c2_random[, 1] <- c(mean_cov_c_random_arm2)
table_c2_random[, 2] <- c(sd_cov_c_random_arm2)
table_c2_random[, 3] <- c(ql_cov_c_random_arm2)
table_c2_random[, 4] <- c(qu_cov_c_random_arm2)
table_c2_random <- round(table_c2_random, digits = digits)
table_list_random$Comparator$Costs <- table_c1_random
table_list_random$Reference$Costs <- table_c2_random
}
if(random == TRUE & p_c_random == 0 & p_e_random == 0) {
stop("No random effects estimates found")
}
}
if(object$data_format == "long") {
max_time <- dim(object$data_set$effects$Control)[2]
time_dep <- object$time_dep
if(time < 1 | time > max_time | is.numeric(time) == FALSE | isTRUE(all.equal(time, as.integer(time))) == FALSE) {
stop("Time must be a numeric integer. Minimum time is 1 (baseline) and maximum time is the latest follow-up time in the data")
}
if(length(grep("^SELECTION", object$model_output$type)) == 1) {
cov_u_fixed <- names(object$data_set$covariates_effects_fixed$Control)
p_u_fixed <- length(cov_u_fixed)
if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 3) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 4) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_effects_fixed)) == 0) {
if(length(dim(object$model_output$covariate_parameter_effects_fixed$alpha)) == 3 & object$model_output$ind_fixed == FALSE) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_fixed$alpha)) == 4 & object$model_output$ind_fixed == FALSE) {
mean_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), mean)
sd_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), sd)
ql_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha[, , , time], c(2, 3), quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_dep_utime_u_fixed <- NULL
dep_dep_ctime_u_fixed <- NULL
if(object$model_output$ind_time_fixed == FALSE) {
if(time >= 2 & time_dep == "AR1") {
dep_dep_utime_u_fixed <- "u(j-1)"
mean_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, mean)
sd_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, sd)
ql_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_u_fixed <- "c(j-1)"
mean_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, mean)
sd_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, sd)
ql_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_u_fixed <- apply(object$model_output$covariate_parameter_effects_fixed$alpha_tc[, , time], 2, quantile, prob = prob[2])
}
}
} else {
dep_dep_utime_u_fixed <- NULL
dep_dep_ctime_u_fixed <- NULL
}
cov_c_fixed <- names(object$data_set$covariates_costs_fixed$Control)
p_c_fixed <- length(cov_c_fixed)
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 3 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 4 & object$model_output$ind_fixed == TRUE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_fixed)) == 0) {
if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 3 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_fixed$beta)) == 4 & object$model_output$ind_fixed == FALSE) {
mean_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), mean)
sd_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), sd)
ql_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta[, , , time], c(2, 3), quantile, prob = prob[2])
}
}
if(object$model_output$ind_fixed == FALSE) {
dep_c_fixed <- "u"
dep_dep_utime_c_fixed <- NULL
dep_dep_ctime_c_fixed <- NULL
mean_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, mean)
sd_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, sd)
ql_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, quantile, prob = prob[1])
qu_dep_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_f[, , time], 2, quantile, prob = prob[2])
if(object$model_output$ind_time_fixed == FALSE) {
if(time >= 2 & time_dep == "AR1") {
dep_dep_utime_c_fixed <- "u(j-1)"
mean_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, mean)
sd_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, sd)
ql_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_c_fixed <- "c(j-1)"
mean_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, mean)
sd_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, sd)
ql_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_c_fixed <- apply(object$model_output$covariate_parameter_costs_fixed$beta_tc[, , time], 2, quantile, prob = prob[2])
}
}
} else {
dep_c_fixed <- NULL
dep_dep_utime_c_fixed <- NULL
dep_dep_ctime_c_fixed <- NULL
}
cov_u_random <- names(object$data_set$covariates_effects_random$Control)
p_u_random <- length(cov_u_random)
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 4) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3), quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_effects_random)) == 0 & is.null(object$model_output$covariate_parameter_effects_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 3) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , time], 2, quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_effects_random$a1)) == 4) {
mean_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , mean)
sd_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , sd)
ql_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , quantile, prob = prob[1])
qu_cov_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1[, , , time], c(2, 3) , quantile, prob = prob[2])
mean_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , mean)
sd_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , sd)
ql_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , quantile, prob = prob[1])
qu_cov_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2[, , , time], c(2, 3) , quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_dep_utime_u_random <- NULL
dep_dep_ctime_u_random <- NULL
if(is.null(object$model_output$covariate_parameter_effects_random$a1_tu) == FALSE) {
if(time >= 2 & time_dep == "AR1") {
dep_dep_utime_u_random <- "u(j-1)"
mean_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, mean)
sd_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, sd)
ql_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tu[, , time], 2, quantile, prob = prob[2])
mean_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, mean)
sd_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, sd)
ql_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_u_random <- "c(j-1)"
mean_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, mean)
sd_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, sd)
ql_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_u1_random <- apply(object$model_output$covariate_parameter_effects_random$a1_tc[, , time], 2, quantile, prob = prob[2])
mean_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, mean)
sd_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, sd)
ql_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_u2_random <- apply(object$model_output$covariate_parameter_effects_random$a2_tc[, , time], 2, quantile, prob = prob[2])
}
} else if(is.null(object$model_output$covariate_parameter_costs_random$a1_tu) == TRUE) {
dep_dep_utime_u_random <- NULL
dep_dep_ctime_u_random <- NULL
}
} else {
dep_dep_utime_u_random <- NULL
dep_dep_ctime_u_random <- NULL
}
cov_c_random <- names(object$data_set$covariates_costs_random$Control)
p_c_random <- length(cov_c_random)
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 3) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random)) == 4) {
mean_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), mean)
sd_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), sd)
ql_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c_random <- apply(object$model_output$covariate_parameter_costs_random[, , , time], c(2, 3), quantile, prob = prob[2])
}
if(length(dim(object$model_output$covariate_parameter_costs_random)) == 0 & is.null(object$model_output$covariate_parameter_costs_random) == FALSE) {
if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 3) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , time], 2, quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , time], 2, quantile, prob = prob[2])
} else if(length(dim(object$model_output$covariate_parameter_costs_random$b1)) == 4) {
mean_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), mean)
sd_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), sd)
ql_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1[, , , time], c(2, 3), quantile, prob = prob[2])
mean_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), mean)
sd_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), sd)
ql_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), quantile, prob = prob[1])
qu_cov_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2[, , , time], c(2, 3), quantile, prob = prob[2])
}
}
if(object$model_output$ind_random == FALSE) {
dep_c_random <- "u"
mean_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, mean)
sd_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, sd)
ql_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, quantile, prob = prob[1])
qu_dep_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_f[, , time], 2, quantile, prob = prob[2])
mean_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, mean)
sd_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, sd)
ql_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, quantile, prob = prob[1])
qu_dep_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_f[, , time], 2, quantile, prob = prob[2])
dep_dep_utime_c_random <- NULL
dep_dep_ctime_c_random <- NULL
if(object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == FALSE) {
if(time >= 2) {
dep_dep_utime_c_random <- "u(j-1)"
mean_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, mean)
sd_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, sd)
ql_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tu[, , time], 2, quantile, prob = prob[2])
mean_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, mean)
sd_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, sd)
ql_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, quantile, prob = prob[1])
qu_dep_utime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tu[, , time], 2, quantile, prob = prob[2])
dep_dep_ctime_c_random <- "c(j-1)"
mean_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, mean)
sd_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, sd)
ql_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_c1_random <- apply(object$model_output$covariate_parameter_costs_random$b1_tc[, , time], 2, quantile, prob = prob[2])
mean_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, mean)
sd_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, sd)
ql_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, quantile, prob = prob[1])
qu_dep_ctime_c2_random <- apply(object$model_output$covariate_parameter_costs_random$b2_tc[, , time], 2, quantile, prob = prob[2])
}
} else if(object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == TRUE) {
dep_dep_utime_c_random <- NULL
dep_dep_ctime_c_random <- NULL
}
} else {
dep_c_random <- NULL
dep_dep_utime_c_random <- NULL
dep_dep_ctime_c_random <- NULL
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_c_fixed != 0) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_c_fixed == 0) {
cov_c_fixed <- c(dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed)
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE & p_c_fixed != 0) {
if(time == 1) {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
} else if(time > 1) {
if(time_dep == "AR1") {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed, dep_dep_utime_c_fixed, dep_dep_ctime_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed, mean_dep_utime_c_fixed, mean_dep_ctime_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed, sd_dep_utime_c_fixed, sd_dep_ctime_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed, ql_dep_utime_c_fixed, ql_dep_ctime_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed, qu_dep_utime_c_fixed, qu_dep_ctime_c_fixed)
} else {
cov_c_fixed <- c(cov_c_fixed, dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_cov_c_fixed, mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_cov_c_fixed, sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_cov_c_fixed, ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_cov_c_fixed, qu_dep_c_fixed)
}
}
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE & p_c_fixed == 0) {
if(time == 1) {
cov_c_fixed <- c(dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed)
} else if(time > 1) {
if(time_dep == "AR1") {
cov_c_fixed <- c(dep_c_fixed, dep_dep_utime_c_fixed, dep_dep_ctime_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed, mean_dep_utime_c_fixed, mean_dep_ctime_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed, sd_dep_utime_c_fixed, sd_dep_ctime_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed, ql_dep_utime_c_fixed, ql_dep_ctime_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed, qu_dep_utime_c_fixed, qu_dep_ctime_c_fixed)
} else {
cov_c_fixed <- c(dep_c_fixed)
p_c_fixed <- length(cov_c_fixed)
mean_cov_c_fixed <- rbind(mean_dep_c_fixed)
sd_cov_c_fixed <- rbind(sd_dep_c_fixed)
ql_cov_c_fixed <- rbind(ql_dep_c_fixed)
qu_cov_c_fixed <- rbind(qu_dep_c_fixed)
}
}
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE) {
if(time == 1) {
cov_u_fixed <- cov_u_fixed
p_u_fixed <- length(cov_u_fixed)
mean_cov_u_fixed <- mean_cov_u_fixed
sd_cov_u_fixed <- sd_cov_u_fixed
ql_cov_u_fixed <- ql_cov_u_fixed
qu_cov_u_fixed <- qu_cov_u_fixed
} else if(time > 1) {
if(time_dep == "AR1") {
cov_u_fixed <- c(cov_u_fixed, dep_dep_utime_u_fixed, dep_dep_ctime_u_fixed)
p_u_fixed <- length(cov_u_fixed)
mean_cov_u_fixed <- rbind(mean_cov_u_fixed, mean_dep_utime_u_fixed, mean_dep_ctime_u_fixed)
sd_cov_u_fixed <- rbind(sd_cov_u_fixed, sd_dep_utime_u_fixed, sd_dep_ctime_u_fixed)
ql_cov_u_fixed <- rbind(ql_cov_u_fixed, ql_dep_utime_u_fixed, ql_dep_ctime_u_fixed)
qu_cov_u_fixed <- rbind(qu_cov_u_fixed, qu_dep_utime_u_fixed, qu_dep_ctime_u_fixed)
} else {
cov_u_fixed <- cov_u_fixed
p_u_fixed <- length(cov_u_fixed)
mean_cov_u_fixed <- mean_cov_u_fixed
sd_cov_u_fixed <- sd_cov_u_fixed
ql_cov_u_fixed <- ql_cov_u_fixed
qu_cov_u_fixed <- qu_cov_u_fixed
}
}
}
if(object$model_output$ind_fixed == TRUE & p_u_fixed == 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
} else if(object$model_output$ind_fixed == TRUE & p_u_fixed > 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[, 1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[, 1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[, 1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[, 1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[, 2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[, 2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[, 2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[, 2]
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_u_fixed == 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
} else if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == TRUE & p_u_fixed > 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[, 1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[, 1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[, 1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[, 1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[, 2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[, 2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[, 2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[, 2]
}
if(object$model_output$ind_fixed == FALSE & object$model_output$ind_time_fixed == FALSE) {
if(time == 1) {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
} else if(time > 1) {
if(time_dep == "AR1") {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[, 1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[, 1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[, 1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[, 1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[, 2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[, 2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[, 2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[, 2]
} else {
mean_cov_u_fixed_arm1 <- mean_cov_u_fixed[1]
sd_cov_u_fixed_arm1 <- sd_cov_u_fixed[1]
ql_cov_u_fixed_arm1 <- ql_cov_u_fixed[1]
qu_cov_u_fixed_arm1 <- qu_cov_u_fixed[1]
mean_cov_u_fixed_arm2 <- mean_cov_u_fixed[2]
sd_cov_u_fixed_arm2 <- sd_cov_u_fixed[2]
ql_cov_u_fixed_arm2 <- ql_cov_u_fixed[2]
qu_cov_u_fixed_arm2 <- qu_cov_u_fixed[2]
}
}
}
if(object$model_output$ind_fixed == TRUE & p_c_fixed == 1) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[2]
} else if(object$model_output$ind_fixed == TRUE & p_c_fixed > 1) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[, 1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[, 1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[, 1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[, 1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[, 2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[, 2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[, 2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[, 2]
}
if(object$model_output$ind_fixed == FALSE & p_c_fixed != 0) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[, 1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[, 1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[, 1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[, 1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[, 2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[, 2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[, 2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[, 2]
} else if(object$model_output$ind_fixed == FALSE & p_c_fixed == 0) {
mean_cov_c_fixed_arm1 <- mean_cov_c_fixed[1]
sd_cov_c_fixed_arm1 <- sd_cov_c_fixed[1]
ql_cov_c_fixed_arm1 <- ql_cov_c_fixed[1]
qu_cov_c_fixed_arm1 <- qu_cov_c_fixed[1]
mean_cov_c_fixed_arm2 <- mean_cov_c_fixed[2]
sd_cov_c_fixed_arm2 <- sd_cov_c_fixed[2]
ql_cov_c_fixed_arm2 <- ql_cov_c_fixed[2]
qu_cov_c_fixed_arm2 <- qu_cov_c_fixed[2]
}
}
if(length(grep("^SELECTION", object$model_output$type)) == 1) {
table_u1_fixed <- matrix(NA, nrow = p_u_fixed, ncol = 4)
rownames(table_u1_fixed) <- cov_u_fixed
colnames(table_u1_fixed) <- c("mean", "sd", "lower", "upper")
table_u1_fixed[, 1] <- c(mean_cov_u_fixed_arm1)
table_u1_fixed[, 2] <- c(sd_cov_u_fixed_arm1)
table_u1_fixed[, 3] <- c(ql_cov_u_fixed_arm1)
table_u1_fixed[, 4] <- c(qu_cov_u_fixed_arm1)
table_u1_fixed <- round(table_u1_fixed, digits = digits)
table_u2_fixed <- matrix(NA, nrow = p_u_fixed, ncol = 4)
rownames(table_u2_fixed) <- cov_u_fixed
colnames(table_u2_fixed) <- c("mean", "sd", "lower", "upper")
table_u2_fixed[, 1] <- c(mean_cov_u_fixed_arm2)
table_u2_fixed[, 2] <- c(sd_cov_u_fixed_arm2)
table_u2_fixed[, 3] <- c(ql_cov_u_fixed_arm2)
table_u2_fixed[, 4] <- c(qu_cov_u_fixed_arm2)
table_u2_fixed <- round(table_u2_fixed, digits = digits)
table_c1_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c1_fixed) <- cov_c_fixed
colnames(table_c1_fixed) <- c("mean", "sd", "lower", "upper")
table_c1_fixed[, 1] <- c(mean_cov_c_fixed_arm1)
table_c1_fixed[, 2] <- c(sd_cov_c_fixed_arm1)
table_c1_fixed[, 3] <- c(ql_cov_c_fixed_arm1)
table_c1_fixed[, 4] <- c(qu_cov_c_fixed_arm1)
table_c1_fixed <- round(table_c1_fixed, digits = digits)
table_c2_fixed <- matrix(NA, nrow = p_c_fixed, ncol = 4)
rownames(table_c2_fixed) <- cov_c_fixed
colnames(table_c2_fixed) <- c("mean", "sd", "lower", "upper")
table_c2_fixed[, 1] <- c(mean_cov_c_fixed_arm2)
table_c2_fixed[, 2] <- c(sd_cov_c_fixed_arm2)
table_c2_fixed[, 3] <- c(ql_cov_c_fixed_arm2)
table_c2_fixed[, 4] <- c(qu_cov_c_fixed_arm2)
table_c2_fixed <- round(table_c2_fixed, digits = digits)
}
table_list_fixed <- list("Comparator" = list("Effects" = table_u1_fixed, "Costs" = table_c1_fixed),
"Reference" = list("Effects" = table_u2_fixed, "Costs" = table_c2_fixed))
table_list_random <- NULL
table_u1_random <- table_c1_random <- NULL
table_u2_random <- table_c2_random <- NULL
table_list_random <- list("Comparator" = list("Effects" = table_u1_random, "Costs" = table_c1_random),
"Reference" = list("Effects" = table_u2_random, "Costs" = table_c2_random))
clus_u_arm1 <- clus_u_arm2 <- NULL
clus_c_arm1 <- clus_c_arm2 <- NULL
if(p_c_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == TRUE) {
cov_c_random <- c(cov_c_random, dep_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random)
} else if(p_c_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == FALSE) {
if(time == 1) {
cov_c_random <- c(cov_c_random, dep_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random)
} else if(time > 1) {
cov_c_random <- c(cov_c_random, dep_c_random, dep_dep_utime_c_random, dep_dep_ctime_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_cov_c1_random, mean_dep_c1_random, mean_dep_utime_c1_random, mean_dep_ctime_c1_random)
sd_cov_c1_random <- rbind(sd_cov_c1_random, sd_dep_c1_random, sd_dep_utime_c1_random, sd_dep_ctime_c1_random)
ql_cov_c1_random <- rbind(ql_cov_c1_random, ql_dep_c1_random, ql_dep_utime_c1_random, ql_dep_ctime_c1_random)
qu_cov_c1_random <- rbind(qu_cov_c1_random, qu_dep_c1_random, qu_dep_utime_c1_random, qu_dep_ctime_c1_random)
mean_cov_c2_random <- rbind(mean_cov_c2_random, mean_dep_c2_random, mean_dep_utime_c2_random, mean_dep_ctime_c2_random)
sd_cov_c2_random <- rbind(sd_cov_c2_random, sd_dep_c2_random, sd_dep_utime_c2_random, sd_dep_ctime_c2_random)
ql_cov_c2_random <- rbind(ql_cov_c2_random, ql_dep_c2_random, ql_dep_utime_c2_random, ql_dep_ctime_c2_random)
qu_cov_c2_random <- rbind(qu_cov_c2_random, qu_dep_c2_random, qu_dep_utime_c2_random, qu_dep_ctime_c2_random)
}
} else if(p_c_random == 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == TRUE) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
pc_random <- 1
} else if(p_c_random == 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$b1_tu) == FALSE) {
if(is.null(dep_dep_utime_c_random) == FALSE) {
if(time == 1) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
} else if(time > 1) {
cov_c_random <- c(dep_c_random, dep_dep_utime_c_random, dep_dep_ctime_c_random)
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- rbind(mean_dep_c1_random, mean_dep_utime_c1_random, mean_dep_ctime_c1_random)
sd_cov_c1_random <- rbind(sd_dep_c1_random, sd_dep_utime_c1_random, sd_dep_ctime_c1_random)
ql_cov_c1_random <- rbind(ql_dep_c1_random, ql_dep_utime_c1_random, ql_dep_ctime_c1_random)
qu_cov_c1_random <- rbind(qu_dep_c1_random, qu_dep_utime_c1_random, qu_dep_ctime_c1_random)
mean_cov_c2_random <- rbind(mean_dep_c2_random, mean_dep_utime_c2_random, mean_dep_ctime_c2_random)
sd_cov_c2_random <- rbind(sd_dep_c2_random, sd_dep_utime_c2_random, sd_dep_ctime_c2_random)
ql_cov_c2_random <- rbind(ql_dep_c2_random, ql_dep_utime_c2_random, ql_dep_ctime_c2_random)
qu_cov_c2_random <- rbind(qu_dep_c2_random, qu_dep_utime_c2_random, qu_dep_ctime_c2_random)
}
} else if(is.null(dep_dep_utime_c_random) == TRUE) {
if(time == 1) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
} else if(time > 1) {
cov_c_random <- dep_c_random
p_c_random <- length(cov_c_random)
mean_cov_c1_random <- mean_dep_c1_random
sd_cov_c1_random <- sd_dep_c1_random
ql_cov_c1_random <- ql_dep_c1_random
qu_cov_c1_random <- qu_dep_c1_random
mean_cov_c2_random <- mean_dep_c2_random
sd_cov_c2_random <- sd_dep_c2_random
ql_cov_c2_random <- ql_dep_c2_random
qu_cov_c2_random <- qu_dep_c2_random
}
}
}
if(p_u_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$a1_tu) == FALSE) {
if(time > 1 & time_dep == "AR1") {
cov_u_random <- c(cov_u_random, dep_dep_utime_u_random, dep_dep_ctime_u_random)
p_u_random <- length(cov_u_random)
mean_cov_u1_random <- rbind(mean_cov_u1_random, mean_dep_utime_u1_random, mean_dep_ctime_u1_random)
sd_cov_u1_random <- rbind(sd_cov_u1_random, sd_dep_utime_u1_random, sd_dep_ctime_u1_random)
ql_cov_u1_random <- rbind(ql_cov_u1_random, ql_dep_utime_u1_random, ql_dep_ctime_u1_random)
qu_cov_u1_random <- rbind(qu_cov_u1_random, qu_dep_utime_u1_random, qu_dep_ctime_u1_random)
mean_cov_u2_random <- rbind(mean_cov_u2_random, mean_dep_utime_u2_random, mean_dep_ctime_u2_random)
sd_cov_u2_random <- rbind(sd_cov_u2_random, sd_dep_utime_u2_random, sd_dep_ctime_u2_random)
ql_cov_u2_random <- rbind(ql_cov_u2_random, ql_dep_utime_u2_random, ql_dep_ctime_u2_random)
qu_cov_u2_random <- rbind(qu_cov_u2_random, qu_dep_utime_u2_random, qu_dep_ctime_u2_random)
}
} else if (p_u_random != 0 & object$model_output$ind_random == FALSE & is.null(object$model_output$covariate_parameter_costs_random$a1_tu) == TRUE) {
if(time > 1 & time_dep == "AR1") {
cov_u_random <- c(cov_u_random, dep_dep_utime_u_random, dep_dep_ctime_u_random)
p_u_random <- length(cov_u_random)
mean_cov_u1_random <- rbind(mean_cov_u1_random, mean_dep_utime_u1_random, mean_dep_ctime_u1_random)
sd_cov_u1_random <- rbind(sd_cov_u1_random, sd_dep_utime_u1_random, sd_dep_ctime_u1_random)
ql_cov_u1_random <- rbind(ql_cov_u1_random, ql_dep_utime_u1_random, ql_dep_ctime_u1_random)
qu_cov_u1_random <- rbind(qu_cov_u1_random, qu_dep_utime_u1_random, qu_dep_ctime_u1_random)
mean_cov_u2_random <- rbind(mean_cov_u2_random, mean_dep_utime_u2_random, mean_dep_ctime_u2_random)
sd_cov_u2_random <- rbind(sd_cov_u2_random, sd_dep_utime_u2_random, sd_dep_ctime_u2_random)
ql_cov_u2_random <- rbind(ql_cov_u2_random, ql_dep_utime_u2_random, ql_dep_ctime_u2_random)
qu_cov_u2_random <- rbind(qu_cov_u2_random, qu_dep_utime_u2_random, qu_dep_ctime_u2_random)
}
}
if(p_u_random != 0) {
mean_cov_u_random_arm1 <- mean_cov_u1_random
sd_cov_u_random_arm1 <- sd_cov_u1_random
ql_cov_u_random_arm1 <- ql_cov_u1_random
qu_cov_u_random_arm1 <- qu_cov_u1_random
mean_cov_u_random_arm2 <- mean_cov_u2_random
sd_cov_u_random_arm2 <- sd_cov_u2_random
ql_cov_u_random_arm2 <- ql_cov_u2_random
qu_cov_u_random_arm2 <- qu_cov_u2_random
clus_u_arm1 <- max(as.numeric(object$data_set$clus_effects$Control))
clus_u_arm2 <- max(as.numeric(object$data_set$clus_effects$Intervention))
clus_u1_index <- sort(rep(seq(1:clus_u_arm1), p_u_random))
clus_u2_index <- sort(rep(seq(1:clus_u_arm2), p_u_random))
table_u1_random <- matrix(NA, nrow = length(clus_u1_index), ncol = 4)
rownames(table_u1_random) <- paste(rep(cov_u_random, clus_u_arm1), clus_u1_index, sep = " ")
colnames(table_u1_random) <- c("mean", "sd", "lower", "upper")
table_u1_random[, 1] <- c(mean_cov_u_random_arm1)
table_u1_random[, 2] <- c(sd_cov_u_random_arm1)
table_u1_random[, 3] <- c(ql_cov_u_random_arm1)
table_u1_random[, 4] <- c(qu_cov_u_random_arm1)
table_u1_random <- round(table_u1_random, digits = digits)
table_u2_random <- matrix(NA, nrow = length(clus_u2_index), ncol = 4)
rownames(table_u2_random) <- paste(rep(cov_u_random, clus_u_arm2), clus_u2_index, sep = " ")
colnames(table_u2_random) <- c("mean", "sd", "lower", "upper")
table_u2_random[, 1] <- c(mean_cov_u_random_arm2)
table_u2_random[, 2] <- c(sd_cov_u_random_arm2)
table_u2_random[, 3] <- c(ql_cov_u_random_arm2)
table_u2_random[, 4] <- c(qu_cov_u_random_arm2)
table_u2_random <- round(table_u2_random, digits = digits)
table_list_random$Comparator$Effects <- table_u1_random
table_list_random$Reference$Effects <- table_u2_random
}
if(p_c_random != 0 | object$model_output$ind_random == FALSE) {
mean_cov_c_random_arm1 <- mean_cov_c1_random
sd_cov_c_random_arm1 <- sd_cov_c1_random
ql_cov_c_random_arm1 <- ql_cov_c1_random
qu_cov_c_random_arm1 <- qu_cov_c1_random
mean_cov_c_random_arm2 <- mean_cov_c2_random
sd_cov_c_random_arm2 <- sd_cov_c2_random
ql_cov_c_random_arm2 <- ql_cov_c2_random
qu_cov_c_random_arm2 <- qu_cov_c2_random
clus_c_arm1 <- max(as.numeric(object$data_set$clus_costs$Control))
clus_c_arm2 <- max(as.numeric(object$data_set$clus_costs$Intervention))
clus_c1_index <- sort(rep(seq(1:clus_c_arm1), p_c_random))
clus_c2_index <- sort(rep(seq(1:clus_c_arm2), p_c_random))
table_c1_random <- matrix(NA, nrow = length(clus_c1_index), ncol = 4)
rownames(table_c1_random) <- paste(rep(cov_c_random, clus_c_arm1), clus_c1_index, sep = " ")
colnames(table_c1_random) <- c("mean", "sd", "lower", "upper")
table_c1_random[, 1] <- c(mean_cov_c_random_arm1)
table_c1_random[, 2] <- c(sd_cov_c_random_arm1)
table_c1_random[, 3] <- c(ql_cov_c_random_arm1)
table_c1_random[, 4] <- c(qu_cov_c_random_arm1)
table_c1_random <- round(table_c1_random, digits = digits)
table_c2_random <- matrix(NA, nrow = length(clus_c2_index), ncol = 4)
rownames(table_c2_random) <- paste(rep(cov_c_random, clus_c_arm2), clus_c2_index, sep = " ")
colnames(table_c2_random) <- c("mean", "sd", "lower", "upper")
table_c2_random[, 1] <- c(mean_cov_c_random_arm2)
table_c2_random[, 2] <- c(sd_cov_c_random_arm2)
table_c2_random[, 3] <- c(ql_cov_c_random_arm2)
table_c2_random[, 4] <- c(qu_cov_c_random_arm2)
table_c2_random <- round(table_c2_random, digits = digits)
table_list_random$Comparator$Costs <- table_c1_random
table_list_random$Reference$Costs <- table_c2_random
}
if(random == TRUE & p_c_random == 0 & p_u_random == 0) {
stop("No random effects estimates found for effect and cost models")
}
}
if(random == FALSE) {
print(table_list_fixed)
} else if(random == TRUE) {
print(table_list_random)
}
}
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.