Nothing
R/timeroc.R
In jsmodule: 'RStudio' Addins and 'Shiny' Modules for Medical Research
Defines functions
timerocModule2
timerocModule
survIDINRI_helper
timeROC_table
timeROChelper
timerocUI
Documented in
survIDINRI_helper
timeROChelper
timerocModule
timerocModule2
timeROC_table
timerocUI
#' @title timerocUI: shiny module UI for time-dependent roc analysis
#' @description Shiny module UI for time-dependent roc analysis
#' @param id id
#' @return Shiny module UI for time-dependent roc analysis
#' @details Shiny module UI for time-dependent roc analysis
#' @examples
#' library(shiny)
#' library(DT)
#' library(data.table)
#' library(jstable)
#' library(ggplot2)
#' library(timeROC)
#' library(survIDINRI)
#' ui <- fluidPage(
#' sidebarLayout(
#' sidebarPanel(
#' timerocUI("timeroc")
#' ),
#' mainPanel(
#' plotOutput("plot_timeroc"),
#' ggplotdownUI("timeroc"),
#' DTOutput("table_timeroc")
#' )
#' )
#' )
#'
#' server <- function(input, output, session) {
#' data <- reactive(mtcars)
#' data.label <- jstable::mk.lev(mtcars)
#'
#' out_timeroc <- callModule(timerocModule, "timeroc",
#' data = data, data_label = data.label,
#' data_varStruct = NULL
#' )
#'
#' output$plot_timeroc <- renderPlot({
#' print(out_timeroc()$plot)
#' })
#'
#' output$table_timeroc <- renderDT({
#' datatable(out_timeroc()$tb,
#' rownames = F, editable = F, extensions = "Buttons",
#' caption = "ROC results",
#' options = c(jstable::opt.tbreg("roctable"), list(scrollX = TRUE))
#' )
#' })
#' }
#' @rdname timerocUI
#' @export
timerocUI <- function(id) {
# Create a namespace function using the provided id
ns <- NS(id)
tagList(
uiOutput(ns("eventtime")),
uiOutput(ns("indep")),
uiOutput(ns("addmodel")),
uiOutput(ns("time")),
checkboxInput(ns("subcheck"), "Sub-group analysis"),
uiOutput(ns("subvar")),
uiOutput(ns("subval"))
)
}
#' @title timeROChelper: Helper function for timerocModule
#' @description Helper function for timerocModule
#' @param var.event event
#' @param var.time time
#' @param vars.ind independent variable
#' @param t time
#' @param data data
#' @param design.survey survey data, Default: NULL
#' @param id.cluster cluster variable if marginal model, Default: NULL
#' @return timeROC and coxph object
#' @details Helper function for timerocModule
#' @examples
#' # library(survival)
#' # timeROChelper("status", "time", c("age", "sex"), t = 365, data = lung)
#' @seealso
#' \code{\link[survival]{coxph}}
#' \code{\link[survey]{svycoxph}}
#' \code{\link[stats]{predict}}
#' \code{\link[timeROC]{timeROC}}
#' @rdname timeROChelper
#' @importFrom survival coxph Surv
#' @importFrom survey svycoxph
#' @importFrom stats predict
#' @importFrom timeROC timeROC
timeROChelper <- function(var.event, var.time, vars.ind, t, data, design.survey = NULL, id.cluster = NULL) {
data[[var.event]] <- as.numeric(as.vector(data[[var.event]]))
forms <- as.formula(paste0("survival::Surv(", var.time, ",", var.event, ") ~ ", paste(vars.ind, collapse = "+")))
cmodel <- NULL
if (is.null(design.survey)) {
if (!is.null(id.cluster)) {
cmodel <- survival::coxph(forms, data = data, y = T, cluster = get(id.cluster))
} else {
cmodel <- survival::coxph(forms, data = data, y = T)
}
} else {
cmodel <- survey::svycoxph(forms, design = design.survey, y = T)
}
lp <- stats::predict(cmodel, type = "lp")
vec.y <- sapply(cmodel$y, `[[`, 1)
out <- timeROC::timeROC(
T = vec.y[1:(length(vec.y) / 2)],
delta = vec.y[(length(vec.y) / 2 + 1):length(vec.y)],
marker = lp,
cause = 1,
weighting = "marginal",
times = t
)
if (out$AUC[2] < 0.5) {
out <- timeROC::timeROC(
T = vec.y[1:(length(vec.y) / 2)],
delta = vec.y[(length(vec.y) / 2 + 1):length(vec.y)],
marker = -lp,
cause = 1,
weighting = "marginal",
times = t
)
}
## Coxph object
data[[var.event]][data[[var.event]] == 1 & data[[var.time]] > t] <- 0
data[[var.time]][data[[var.time]] > t] <- t
if (is.null(design.survey)) {
if (!is.null(id.cluster)) {
cmodel <- survival::coxph(forms, data = data, y = T, cluster = get(id.cluster))
} else {
cmodel <- survival::coxph(forms, data = data, y = T)
}
} else {
cmodel <- survey::svycoxph(forms, design = design.survey, y = T)
}
return(list(coxph = cmodel, timeROC = out))
}
#' @title timeROC_table: extract AUC information from list of timeROChelper object.
#' @description extract AUC information from list of timeROChelper object.
#' @param ListModel list of timeROChelper object
#' @param dec.auc digits for AUC, Default: 3
#' @param dec.p digits for p value, Default: 3
#' @return table of AUC information
#' @details extract AUC information from list of timeROChelper object.
#' @examples
#' # library(survival)
#' # list.timeROC <- lapply(list("age", c("age", "sex")),
#' # function(x){
#' # timeROChelper("status", "time", x, t = 365, data = lung)
#' # })
#' # timeROC_table(list.timeROC)
#' @seealso
#' \code{\link[stats]{confint}}
#' \code{\link[data.table]{data.table}}
#' @rdname timeROC_table
#' @importFrom stats confint qnorm
#' @importFrom data.table data.table
#' @importFrom survival concordance
timeROC_table <- function(ListModel, dec.auc = 3, dec.p = 3) {
res.roc <- eval(parse(text = paste0("survival::concordance(", paste(paste0("lapply(ListModel, `[[`, 'coxph')[[", seq_along(ListModel), "]]"), collapse = ", "), ")")))
auc <- res.roc$concordance
se1.96 <- stats::qnorm(0.975) * sqrt(ifelse(length(ListModel) == 1, res.roc$var, diag(res.roc$var)))
auc.ci <- paste0(round(auc - se1.96, dec.auc), "-", round(auc + se1.96, dec.auc))
auc <- round(auc, dec.auc)
if (length(ListModel) == 1) {
out <- data.table::data.table(paste0("Model ", seq_along(ListModel)), auc, auc.ci)
names(out) <- c("Prediction Model", "AUC", "95% CI")
} else {
auc.pdiff <- c(NA, sapply(
seq_along(ListModel)[-1],
function(x) {
contr <- c(-1, 1)
dtest <- contr %*% res.roc$concordance[(x - 1):x]
dvar <- contr %*% res.roc$var[(x - 1):x, (x - 1):x] %*% contr
p <- 2 * pnorm(abs(dtest / sqrt(dvar)), lower.tail = F)
p <- ifelse(p < 0.001, "< 0.001", round(p, dec.p))
return(p)
}
))
out <- data.table::data.table(paste0("Model ", seq_along(ListModel)), auc, auc.ci, auc.pdiff)
names(out) <- c("Prediction Model", "AUC", "95% CI", "P-value for AUC Difference")
}
return(out[])
}
#' @title survIDINRI_helper: Helper function for IDI.INF.OUT in survIDINRI packages
#' @description Helper function for IDI.INF.OUT in survIDINRI packages
#' @param var.event event
#' @param var.time time
#' @param list.vars.ind list of independent variable
#' @param t time
#' @param data data
#' @param dec.auc digits for AUC, Default: 3
#' @param dec.p digits for p value, Default: 3
#' @param id.cluster cluster variable if marginal model, Default: NULL
#' @return IDI, NRI
#' @details Helper function for IDI.INF.OUT in survIDINRI packages
#' @examples
#' # library(survival)
#' # survIDINRI_helper("status", "time", list.vars.ind = list("age", c("age", "sex")),
#' # t = 365, data = lung)
#' @seealso
#' \code{\link[data.table]{data.table}}
#' \code{\link[stats]{model.matrix}}
#' \code{\link[survival]{coxph}}
#' \code{\link[survival]{Surv}}
#' \code{\link[survIDINRI]{IDI.INF.OUT}}
#' \code{\link[survIDINRI]{IDI.INF}}
#' @rdname survIDINRI_helper
#' @importFrom data.table data.table
#' @importFrom stats model.matrix
#' @importFrom survival coxph Surv
#' @importFrom survIDINRI IDI.INF.OUT IDI.INF
survIDINRI_helper <- function(var.event, var.time, list.vars.ind, t, data, dec.auc = 3, dec.p = 3, id.cluster = NULL) {
data <- data.table::data.table(data)
data[[var.event]] <- as.numeric(as.vector(data[[var.event]]))
vars <- c(Reduce(union, list(var.event, var.time, unlist(list.vars.ind))))
if (!is.null(id.cluster)) {
data <- na.omit(data[, .SD, .SDcols = c(vars, id.cluster)])
} else {
data <- na.omit(data[, .SD, .SDcols = vars])
}
mm <- lapply(
list.vars.ind,
function(x) {
if (!is.null(id.cluster)) {
stats::model.matrix(survival::coxph(as.formula(paste0("survival::Surv(", var.time, ",", var.event, ") ~ ", paste(x, collapse = "+"), "+ cluster(", id.cluster, ")")), data = data))
} else {
stats::model.matrix(survival::coxph(as.formula(paste0("survival::Surv(", var.time, ",", var.event, ") ~ ", paste(x, collapse = "+"))), data = data))
}
}
)
res <- lapply(
seq_along(list.vars.ind)[-1],
function(x) {
resIDINRI <- survIDINRI::IDI.INF.OUT(survIDINRI::IDI.INF(data[, .SD, .SDcols = c(var.time, var.event)], mm[[x - 1]], mm[[x]], t, npert = 200))
zz <- lapply(
list(resIDINRI[1, ], resIDINRI[2, ]),
function(x) {
c(round(x[1], dec.auc), paste0(round(x[2], dec.auc), "-", round(x[3], dec.auc)), ifelse(x[4] < 0.001, "< 0.001", round(x[4], dec.p)))
}
)
return(unlist(zz))
}
)
out <- data.table::data.table(Reduce(rbind, c(list(rep(NA, 6)), res)))
names(out) <- c("IDI", "95% CI", "P-value for IDI", "continuous NRI", "95% CI", "P-value for NRI")
return(out[])
}
#' @title timerocModule: shiny module server for time-dependent roc analysis
#' @description shiny module server for time-dependent roc analysis
#' @param input input
#' @param output output
#' @param session session
#' @param data Reactive data
#' @param data_label Reactuve data label
#' @param data_varStruct Reactive List of variable structure, Default: NULL
#' @param nfactor.limit nlevels limit in factor variable, Default: 10
#' @param design.survey Reactive survey data. default: NULL
#' @param id.cluster Reactive cluster variable if marginal model, Default: NULL
#' @param iid logical, get CI of AUC, Default: T
#' @param NRIIDI logical, get NRI & IDI, Default: T
#' @return shiny module server for time-dependent roc analysis
#' @details shiny module server for time-dependent roc analysis
#' @examples
#' library(shiny)
#' library(DT)
#' library(data.table)
#' library(jstable)
#' library(ggplot2)
#' library(timeROC)
#' library(survIDINRI)
#' ui <- fluidPage(
#' sidebarLayout(
#' sidebarPanel(
#' timerocUI("timeroc")
#' ),
#' mainPanel(
#' plotOutput("plot_timeroc"),
#' ggplotdownUI("timeroc"),
#' DTOutput("table_timeroc")
#' )
#' )
#' )
#'
#' server <- function(input, output, session) {
#' data <- reactive(mtcars)
#' data.label <- jstable::mk.lev(mtcars)
#'
#' out_timeroc <- callModule(timerocModule, "timeroc",
#' data = data, data_label = data.label,
#' data_varStruct = NULL
#' )
#'
#' output$plot_timeroc <- renderPlot({
#' print(out_timeroc()$plot)
#' })
#'
#' output$table_timeroc <- renderDT({
#' datatable(out_timeroc()$tb,
#' rownames = F, editable = F, extensions = "Buttons",
#' caption = "ROC results",
#' options = c(jstable::opt.tbreg("roctable"), list(scrollX = TRUE))
#' )
#' })
#' }
#' @seealso
#' \code{\link[stats]{quantile}}
#' \code{\link[data.table]{setkey}}
#' \code{\link[data.table]{data.table}}
#' \code{\link[data.table]{rbindlist}}
#' @rdname timerocModule
#' @export
#' @importFrom stats quantile median
#' @importFrom data.table setkey rbindlist data.table
#' @importFrom rvg dml
#' @importFrom officer read_pptx add_slide ph_with ph_location
#' @importFrom timeROC SeSpPPVNPV
timerocModule <- function(input, output, session, data, data_label, data_varStruct = NULL, nfactor.limit = 10, design.survey = NULL, id.cluster = NULL, iid = T, NRIIDI = T) {
## To remove NOTE.
ListModel <- compare <- level <- variable <- FP <- TP <- model <- Sensitivity <- Specificity <- NULL
if (is.null(data_varStruct)) {
data_varStruct <- reactive(list(variable = names(data())))
}
vlist <- reactive({
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
factor_vars <- names(data())[data()[, lapply(.SD, class) %in% c("factor", "character")]]
# factor_vars <- names(data())[sapply(names(data()), function(x){class(data()[[x]]) %in% c("factor", "character")})]
factor_list <- mklist(data_varStruct(), factor_vars)
conti_vars <- setdiff(names(data()), factor_vars)
if (!is.null(design.survey)) {
conti_vars <- setdiff(conti_vars, c(names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
}
conti_vars_positive <- conti_vars[unlist(data()[, lapply(.SD, function(x) {
min(x, na.rm = T) >= 0
}), .SDcols = conti_vars])]
conti_list <- mklist(data_varStruct(), conti_vars)
nclass_factor <- unlist(data()[, lapply(.SD, function(x) {
length(levels(x))
}), .SDcols = factor_vars])
# nclass_factor <- sapply(factor_vars, function(x){length(unique(data()[[x]]))})
class01_factor <- unlist(data()[, lapply(.SD, function(x) {
identical(levels(x), c("0", "1"))
}), .SDcols = factor_vars])
validate(
need(length(class01_factor) >= 1, "No categorical variables coded as 0, 1 in data")
)
factor_01vars <- factor_vars[class01_factor]
factor_01_list <- mklist(data_varStruct(), factor_01vars)
group_vars <- factor_vars[nclass_factor >= 2 & nclass_factor <= nfactor.limit & nclass_factor < nrow(data())]
group_list <- mklist(data_varStruct(), group_vars)
except_vars <- factor_vars[nclass_factor > nfactor.limit | nclass_factor == 1 | nclass_factor == nrow(data())]
return(list(
factor_vars = factor_vars, factor_list = factor_list, conti_vars = conti_vars, conti_list = conti_list, conti_vars_positive = conti_vars_positive,
factor_01vars = factor_01vars, factor_01_list = factor_01_list, group_vars = group_vars, group_list = group_list, except_vars = except_vars
))
})
output$eventtime <- renderUI({
validate(
need(length(vlist()$factor_01vars) >= 1, "No candidate event variables coded as 0, 1"),
need(length(vlist()$conti_vars_positive) >= 1, "No candidate time variables")
)
tagList(
selectInput(session$ns("event_km"), "Event",
choices = mklist(data_varStruct(), vlist()$factor_01vars), multiple = F,
selected = NULL
),
selectInput(session$ns("time_km"), "Time",
choices = mklist(data_varStruct(), vlist()$conti_vars_positive), multiple = F,
selected = NULL
)
)
})
nmodel <- reactiveVal(1)
output$addmodel <- renderUI({
if (nmodel() <= 1) {
actionButton(session$ns("add"), label = "Add model", icon("plus"), class = "btn-primary")
} else if (nmodel() > 1) {
tagList(
actionButton(session$ns("add"), label = "Add model", icon("plus"), class = "btn-primary"),
actionButton(session$ns("rmv"), label = "Remove model", icon("minus"))
)
}
})
indeproc <- reactive({
req(!is.null(input$event_km))
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
if (!is.null(design.survey)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
} else if (!is.null(id.cluster)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, id.cluster()))
} else {
indep.roc <- setdiff(names(data()), c(vlist()$except_vars, input$event_km))
}
return(indep.roc)
})
output$indep <- renderUI({
selectInput(session$ns(paste0("indep_km", 1)), paste0("Independent variables for Model ", 1),
choices = mklist(data_varStruct(), indeproc()), multiple = T,
selected = unlist(mklist(data_varStruct(), indeproc()))[1]
)
})
observeEvent(input$add, {
insertUI(
selector = paste0("div:has(> #", session$ns("add"), ")"),
where = "beforeBegin",
ui = selectInput(session$ns(paste0("indep_km", nmodel() + 1)), paste0("Independent variables for Model ", nmodel() + 1),
choices = mklist(data_varStruct(), indeproc()), multiple = T,
selected = unlist(mklist(data_varStruct(), indeproc()))[1:min(length(indeproc()), nmodel() + 1)]
)
)
nmodel(nmodel() + 1)
})
observeEvent(input$rmv, {
removeUI(
selector = paste0("div:has(>> #", session$ns(paste0("indep_km", nmodel())), ")")
)
nmodel(nmodel() - 1)
})
indeps <- reactive(lapply(1:nmodel(), function(i) {
input[[paste0("indep_km", i)]]
}))
output$time <- renderUI({
req(input$time_km)
tvar <- data()[[input$time_km]]
if (min(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
} else if (max(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 365, step = 5)
} else if (max(tvar, na.rm = T) >= 12) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 12)
} else {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
}
})
observeEvent(input$subcheck, {
output$subvar <- renderUI({
req(input$subcheck == T)
indeps.unique <- unique(unlist(indeps()))
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique))
if (!is.null(id.cluster)) {
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique, id.cluster()))
} else if (!is.null(design.survey)) {
var_subgroup <- setdiff(names(data()), union(c(names(design.survey()$strata), names(design.survey()$cluster), names(design.survey()$allprob)), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique)))
}
var_subgroup_list <- mklist(data_varStruct(), var_subgroup)
validate(
need(length(var_subgroup) > 0, "No variables for sub-group analysis")
)
tagList(
selectInput(session$ns("subvar_km"), "Sub-group variables",
choices = var_subgroup_list, multiple = T,
selected = var_subgroup[1]
)
)
})
})
output$subval <- renderUI({
req(input$subcheck == T)
req(length(input$subvar_km) > 0)
outUI <- tagList()
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
outUI[[v]] <- selectInput(session$ns(paste0("subval_km", v)), paste0("Sub-group value: ", input$subvar_km[[v]]),
choices = data_label()[variable == input$subvar_km[[v]], level], multiple = T,
selected = data_label()[variable == input$subvar_km[[v]], level][1]
)
} else {
val <- stats::quantile(data()[[input$subvar_km[[v]]]], na.rm = T)
outUI[[v]] <- sliderInput(session$ns(paste0("subval_km", v)), paste0("Sub-group range: ", input$subvar_km[[v]]),
min = val[1], max = val[5],
value = c(val[2], val[4])
)
}
}
outUI
})
timerocList <- reactive({
req(!is.null(input$event_km))
req(!is.null(input$time_km))
# req(!is.null(input$indep_km1))
# req(!is.null(input$indep_km2))
for (i in 1:nmodel()) {
req(!is.null(input[[paste0("indep_km", i)]]))
}
req(!is.null(indeps()))
collapse.indep <- sapply(1:nmodel(), function(i) {
paste0(input[[paste0("indep_km", i)]], collapse = "")
})
validate(
need(anyDuplicated(collapse.indep) == 0, "Please select different models")
)
data.km <- data()[complete.cases(data()[, .SD, .SDcols = unique(unlist(indeps()))])]
label.regress <- data_label()
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.km <- data.km[get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]]]
} else {
data.km <- data.km[get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2]]
}
}
data.km[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.km)[, c("variable", "level")]
data.table::setkey(data_label(), "variable", "level")
data.table::setkey(label.regress2, "variable", "level")
label.regress <- data_label()[label.regress2]
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
}
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km)
})
if ((nmodel() == 1 | NRIIDI == F)) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
res.cut <- NULL
}
} else {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km, id.cluster = id.cluster())
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km, id.cluster = id.cluster()
)
)
res.cut <- NULL
}
}
# res.tb <- timeROC_table(res.roc)
} else {
data.design <- design.survey()
label.regress <- data_label()
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.design <- subset(data.design, get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]])
} else {
data.design <- subset(data.design, get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2])
}
}
data.design$variables[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.design$variables)[, c("variable", "class", "level")]
data.table::setkey(data_label(), "variable", "class", "level")
data.table::setkey(label.regress2, "variable", "class", "level")
label.regress <- data_label()[label.regress2]
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
}
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km,
vars.ind = x,
t = input$time_to_roc, data = data.km, design.survey = data.design
)
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
res.cut <- NULL
}
}
res.timeROC <- lapply(res.roc, `[[`, "timeROC")
data.rocplot <- data.table::rbindlist(
lapply(
1:length(res.timeROC),
function(x) {
data.table::data.table(
FP = res.timeROC[[x]]$FP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
TP = res.timeROC[[x]]$TP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
model = paste0("model ", x)
)
}
)
)
p <- ggplot(data.rocplot, aes(FP, TP, colour = model)) +
geom_line() +
geom_abline(slope = 1, lty = 2) +
xlab("1-Specificity") +
ylab("Sensitivity")
return(list(plot = p, tb = res.tb, cut = res.cut))
})
output$downloadControls <- renderUI({
tagList(
column(
4,
selectizeInput(session$ns("file_ext"), "File extension (dpi = 300)",
choices = c("jpg", "pdf", "tiff", "svg", "pptx"), multiple = F,
selected = "pptx"
)
),
column(
4,
sliderInput(session$ns("fig_width"), "Width (in):",
min = 5, max = 15, value = 8
)
),
column(
4,
sliderInput(session$ns("fig_height"), "Height (in):",
min = 5, max = 15, value = 6
)
)
)
})
output$downloadButton <- downloadHandler(
filename = function() {
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
return(paste(input$event_km, "_", input$time_km, "_timeROC.", input$file_ext, sep = ""))
} else {
return(paste(input$event_km, "_", input$time_km, "_timeROC_marginal.", input$file_ext, sep = ""))
}
} else {
return(paste(input$event_km, "_", input$time_km, "__timeROC_survey.", input$file_ext, sep = ""))
}
},
# content is a function with argument file. content writes the plot to the device
content = function(file) {
withProgress(
message = "Download in progress",
detail = "This may take a while...",
value = 0,
{
for (i in 1:15) {
incProgress(1 / 15)
Sys.sleep(0.01)
}
if (input$file_ext == "pptx") {
my_vec_graph <- rvg::dml(ggobj = timerocList()$plot)
doc <- officer::read_pptx()
doc <- officer::add_slide(doc, layout = "Title and Content", master = "Office Theme")
doc <- officer::ph_with(doc, my_vec_graph, location = officer::ph_location(width = input$fig_width, height = input$fig_height))
print(doc, target = file)
} else {
ggsave(file, timerocList()$plot, dpi = 300, units = "in", width = input$fig_width, height = input$fig_height)
}
}
)
}
)
return(timerocList)
}
#' @title timerocModule2: shiny module server for time dependent roc analysis- input number of model as integer
#' @description shiny module server for time-dependent roc analysis- input number of model as integer
#' @param input input
#' @param output output
#' @param session session
#' @param data Reactive data
#' @param data_label Reactuve data label
#' @param data_varStruct Reactive List of variable structure, Default: NULL
#' @param nfactor.limit nlevels limit in factor variable, Default: 10
#' @param design.survey Reactive survey data. default: NULL
#' @param id.cluster Reactive cluster variable if marginal model, Default: NULL
#' @param iid logical, get CI of AUC, Default: T
#' @param NRIIDI logical, get NRI & IDI, Default: T
#' @return shiny module server for time dependent roc analysis- input number of model as integer
#' @details shiny module server for time dependent roc analysis- input number of model as integer
#' @examples
#' library(shiny)
#' library(DT)
#' library(data.table)
#' library(jstable)
#' library(ggplot2)
#' library(timeROC)
#' library(survIDINRI)
#' ui <- fluidPage(
#' sidebarLayout(
#' sidebarPanel(
#' timerocUI("timeroc")
#' ),
#' mainPanel(
#' plotOutput("plot_timeroc"),
#' ggplotdownUI("timeroc"),
#' DTOutput("table_timeroc")
#' )
#' )
#' )
#'
#' server <- function(input, output, session) {
#' data <- reactive(mtcars)
#' data.label <- jstable::mk.lev(mtcars)
#'
#' out_timeroc <- callModule(timerocModule2, "timeroc",
#' data = data, data_label = data.label,
#' data_varStruct = NULL
#' )
#'
#' output$plot_timeroc <- renderPlot({
#' print(out_timeroc()$plot)
#' })
#'
#' output$table_timeroc <- renderDT({
#' datatable(out_timeroc()$tb,
#' rownames = F, editable = F, extensions = "Buttons",
#' caption = "ROC results",
#' options = c(jstable::opt.tbreg("roctable"), list(scrollX = TRUE))
#' )
#' })
#' }
#' @seealso
#' \code{\link[stats]{quantile}}
#' \code{\link[data.table]{setkey}}
#' \code{\link[data.table]{data.table}}
#' \code{\link[data.table]{rbindlist}}
#' @rdname timerocModule
#' @export
#' @importFrom stats quantile median
#' @importFrom data.table setkey rbindlist data.table
#' @importFrom rvg dml
#' @importFrom officer read_pptx add_slide ph_with ph_location
timerocModule2 <- function(input, output, session, data, data_label, data_varStruct = NULL, nfactor.limit = 10, design.survey = NULL, id.cluster = NULL, iid = T, NRIIDI = T) {
## To remove NOTE.
ListModel <- compare <- level <- variable <- FP <- TP <- model <- Sensitivity <- Specificity <- NULL
if (is.null(data_varStruct)) {
data_varStruct <- reactive(list(variable = names(data())))
}
vlist <- reactive({
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
factor_vars <- names(data())[data()[, lapply(.SD, class) %in% c("factor", "character")]]
# factor_vars <- names(data())[sapply(names(data()), function(x){class(data()[[x]]) %in% c("factor", "character")})]
factor_list <- mklist(data_varStruct(), factor_vars)
conti_vars <- setdiff(names(data()), factor_vars)
if (!is.null(design.survey)) {
conti_vars <- setdiff(conti_vars, c(names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
}
conti_vars_positive <- conti_vars[unlist(data()[, lapply(.SD, function(x) {
min(x, na.rm = T) >= 0
}), .SDcols = conti_vars])]
conti_list <- mklist(data_varStruct(), conti_vars)
nclass_factor <- unlist(data()[, lapply(.SD, function(x) {
length(levels(x))
}), .SDcols = factor_vars])
# nclass_factor <- sapply(factor_vars, function(x){length(unique(data()[[x]]))})
class01_factor <- unlist(data()[, lapply(.SD, function(x) {
identical(levels(x), c("0", "1"))
}), .SDcols = factor_vars])
validate(
need(length(class01_factor) >= 1, "No categorical variables coded as 0, 1 in data")
)
factor_01vars <- factor_vars[class01_factor]
factor_01_list <- mklist(data_varStruct(), factor_01vars)
group_vars <- factor_vars[nclass_factor >= 2 & nclass_factor <= nfactor.limit & nclass_factor < nrow(data())]
group_list <- mklist(data_varStruct(), group_vars)
except_vars <- factor_vars[nclass_factor > nfactor.limit | nclass_factor == 1 | nclass_factor == nrow(data())]
return(list(
factor_vars = factor_vars, factor_list = factor_list, conti_vars = conti_vars, conti_list = conti_list, conti_vars_positive = conti_vars_positive,
factor_01vars = factor_01vars, factor_01_list = factor_01_list, group_vars = group_vars, group_list = group_list, except_vars = except_vars
))
})
output$eventtime <- renderUI({
validate(
need(length(vlist()$factor_01vars) >= 1, "No candidate event variables coded as 0, 1"),
need(length(vlist()$conti_vars_positive) >= 1, "No candidate time variables")
)
tagList(
selectInput(session$ns("event_km"), "Event",
choices = mklist(data_varStruct(), vlist()$factor_01vars), multiple = F,
selected = NULL
),
selectInput(session$ns("time_km"), "Time",
choices = mklist(data_varStruct(), vlist()$conti_vars_positive), multiple = F,
selected = NULL
)
)
})
output$addmodel <- renderUI({
radioButtons(session$ns("nmodel"), "Number of models", 1:5, selected = 1, inline = T)
})
nmodel <- reactive(as.integer(input$nmodel))
indeproc <- reactive({
req(!is.null(input$event_km))
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
if (!is.null(design.survey)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
} else if (!is.null(id.cluster)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, id.cluster()))
} else {
indep.roc <- setdiff(names(data()), c(vlist()$except_vars, input$event_km))
}
return(indep.roc)
})
output$indep <- renderUI({
req(nmodel())
lapply(1:nmodel(), function(x) {
selectInput(session$ns(paste0("indep_km", x)), paste0("Independent variables for Model ", x),
choices = mklist(data_varStruct(), indeproc()), multiple = T,
selected = unlist(mklist(data_varStruct(), indeproc()))[x]
)
})
})
indeps <- reactive(lapply(1:nmodel(), function(i) {
input[[paste0("indep_km", i)]]
}))
output$time <- renderUI({
req(input$time_km)
tvar <- data()[[input$time_km]]
if (min(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
} else if (max(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 365, step = 5)
} else if (max(tvar, na.rm = T) >= 12) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 12)
} else {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
}
})
observeEvent(input$subcheck, {
output$subvar <- renderUI({
req(input$subcheck == T)
indeps.unique <- unique(unlist(indeps()))
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique))
if (!is.null(id.cluster)) {
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique, id.cluster()))
} else if (!is.null(design.survey)) {
var_subgroup <- setdiff(names(data()), union(c(names(design.survey()$strata), names(design.survey()$cluster), names(design.survey()$allprob)), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique)))
}
var_subgroup_list <- mklist(data_varStruct(), var_subgroup)
validate(
need(length(var_subgroup) > 0, "No variables for sub-group analysis")
)
tagList(
selectInput(session$ns("subvar_km"), "Sub-group variables",
choices = var_subgroup_list, multiple = T,
selected = var_subgroup[1]
)
)
})
})
output$subval <- renderUI({
req(input$subcheck == T)
req(length(input$subvar_km) > 0)
outUI <- tagList()
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
outUI[[v]] <- selectInput(session$ns(paste0("subval_km", v)), paste0("Sub-group value: ", input$subvar_km[[v]]),
choices = data_label()[variable == input$subvar_km[[v]], level], multiple = T,
selected = data_label()[variable == input$subvar_km[[v]], level][1]
)
} else {
val <- stats::quantile(data()[[input$subvar_km[[v]]]], na.rm = T)
outUI[[v]] <- sliderInput(session$ns(paste0("subval_km", v)), paste0("Sub-group range: ", input$subvar_km[[v]]),
min = val[1], max = val[5],
value = c(val[2], val[4])
)
}
}
outUI
})
timerocList <- reactive({
req(!is.null(input$event_km))
req(!is.null(input$time_km))
# req(!is.null(input$indep_km1))
# req(!is.null(input$indep_km2))
for (i in 1:nmodel()) {
req(!is.null(input[[paste0("indep_km", i)]]))
}
req(!is.null(indeps()))
collapse.indep <- sapply(1:nmodel(), function(i) {
paste0(input[[paste0("indep_km", i)]], collapse = "")
})
validate(
need(anyDuplicated(collapse.indep) == 0, "Please select different models")
)
data.km <- data()
label.regress <- data_label()
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.km <- data.km[get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]]]
} else {
data.km <- data.km[get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2]]
}
}
data.km[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.km)[, c("variable", "class", "level")]
data.table::setkey(data_label(), "variable", "class", "level")
data.table::setkey(label.regress2, "variable", "class", "level")
label.regress <- data_label()[label.regress2]
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
}
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km)
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
}
} else {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km, id.cluster = id.cluster())
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km, id.cluster = id.cluster()
)
)
res.cut <- NULL
}
}
# res.tb <- timeROC_table(res.roc)
} else {
data.design <- design.survey()
label.regress <- data_label()
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.design <- subset(data.design, get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]])
} else {
data.design <- subset(data.design, get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2])
}
}
data.design$variables[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.design$variables)[, c("variable", "class", "level")]
data.table::setkey(data_label(), "variable", "class", "level")
data.table::setkey(label.regress2, "variable", "class", "level")
label.regress <- data_label()[label.regress2]
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
}
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km,
vars.ind = x,
t = input$time_to_roc, data = data.km, design.survey = data.design
)
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
res.cut <- data.table::rbindlist(lapply(unique(data.km[[indeps()[[1]][1]]]), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
res.cut <- NULL
}
}
res.timeROC <- lapply(res.roc, `[[`, "timeROC")
data.rocplot <- data.table::rbindlist(
lapply(
1:length(res.timeROC),
function(x) {
data.table::data.table(
FP = res.timeROC[[x]]$FP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
TP = res.timeROC[[x]]$TP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
model = paste0("model ", x)
)
}
)
)
p <- ggplot(data.rocplot, aes(FP, TP, colour = model)) +
geom_line() +
geom_abline(slope = 1, lty = 2) +
xlab("1-Specificity") +
ylab("Sensitivity")
return(list(plot = p, tb = res.tb))
})
output$downloadControls <- renderUI({
tagList(
column(
4,
selectizeInput(session$ns("file_ext"), "File extension (dpi = 300)",
choices = c("jpg", "pdf", "tiff", "svg", "pptx"), multiple = F,
selected = "pptx"
)
),
column(
4,
sliderInput(session$ns("fig_width"), "Width (in):",
min = 5, max = 15, value = 8
)
),
column(
4,
sliderInput(session$ns("fig_height"), "Height (in):",
min = 5, max = 15, value = 6
)
)
)
})
output$downloadButton <- downloadHandler(
filename = function() {
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
return(paste(input$event_km, "_", input$time_km, "_timeROC.", input$file_ext, sep = ""))
} else {
return(paste(input$event_km, "_", input$time_km, "_timeROC_marginal.", input$file_ext, sep = ""))
}
} else {
return(paste(input$event_km, "_", input$time_km, "__timeROC_survey.", input$file_ext, sep = ""))
}
},
# content is a function with argument file. content writes the plot to the device
content = function(file) {
withProgress(
message = "Download in progress",
detail = "This may take a while...",
value = 0,
{
for (i in 1:15) {
incProgress(1 / 15)
Sys.sleep(0.01)
}
if (input$file_ext == "pptx") {
my_vec_graph <- rvg::dml(ggobj = timerocList()$plot)
doc <- officer::read_pptx()
doc <- officer::add_slide(doc, layout = "Title and Content", master = "Office Theme")
doc <- officer::ph_with(doc, my_vec_graph, location = officer::ph_location(width = input$fig_width, height = input$fig_height))
print(doc, target = file)
} else {
ggsave(file, timerocList()$plot, dpi = 300, units = "in", width = input$fig_width, height = input$fig_height)
}
}
)
}
)
return(timerocList)
}
Try the jsmodule package in your browser
Any scripts or data that you put into this service are public.
jsmodule documentation built on Sept. 11, 2024, 5:30 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 timerocModule2 timerocModule survIDINRI_helper timeROC_table timeROChelper timerocUI
Documented in survIDINRI_helper timeROChelper timerocModule timerocModule2 timeROC_table timerocUI
#' @title timerocUI: shiny module UI for time-dependent roc analysis
#' @description Shiny module UI for time-dependent roc analysis
#' @param id id
#' @return Shiny module UI for time-dependent roc analysis
#' @details Shiny module UI for time-dependent roc analysis
#' @examples
#' library(shiny)
#' library(DT)
#' library(data.table)
#' library(jstable)
#' library(ggplot2)
#' library(timeROC)
#' library(survIDINRI)
#' ui <- fluidPage(
#' sidebarLayout(
#' sidebarPanel(
#' timerocUI("timeroc")
#' ),
#' mainPanel(
#' plotOutput("plot_timeroc"),
#' ggplotdownUI("timeroc"),
#' DTOutput("table_timeroc")
#' )
#' )
#' )
#'
#' server <- function(input, output, session) {
#' data <- reactive(mtcars)
#' data.label <- jstable::mk.lev(mtcars)
#'
#' out_timeroc <- callModule(timerocModule, "timeroc",
#' data = data, data_label = data.label,
#' data_varStruct = NULL
#' )
#'
#' output$plot_timeroc <- renderPlot({
#' print(out_timeroc()$plot)
#' })
#'
#' output$table_timeroc <- renderDT({
#' datatable(out_timeroc()$tb,
#' rownames = F, editable = F, extensions = "Buttons",
#' caption = "ROC results",
#' options = c(jstable::opt.tbreg("roctable"), list(scrollX = TRUE))
#' )
#' })
#' }
#' @rdname timerocUI
#' @export
timerocUI <- function(id) {
# Create a namespace function using the provided id
ns <- NS(id)
tagList(
uiOutput(ns("eventtime")),
uiOutput(ns("indep")),
uiOutput(ns("addmodel")),
uiOutput(ns("time")),
checkboxInput(ns("subcheck"), "Sub-group analysis"),
uiOutput(ns("subvar")),
uiOutput(ns("subval"))
)
}
#' @title timeROChelper: Helper function for timerocModule
#' @description Helper function for timerocModule
#' @param var.event event
#' @param var.time time
#' @param vars.ind independent variable
#' @param t time
#' @param data data
#' @param design.survey survey data, Default: NULL
#' @param id.cluster cluster variable if marginal model, Default: NULL
#' @return timeROC and coxph object
#' @details Helper function for timerocModule
#' @examples
#' # library(survival)
#' # timeROChelper("status", "time", c("age", "sex"), t = 365, data = lung)
#' @seealso
#' \code{\link[survival]{coxph}}
#' \code{\link[survey]{svycoxph}}
#' \code{\link[stats]{predict}}
#' \code{\link[timeROC]{timeROC}}
#' @rdname timeROChelper
#' @importFrom survival coxph Surv
#' @importFrom survey svycoxph
#' @importFrom stats predict
#' @importFrom timeROC timeROC
timeROChelper <- function(var.event, var.time, vars.ind, t, data, design.survey = NULL, id.cluster = NULL) {
data[[var.event]] <- as.numeric(as.vector(data[[var.event]]))
forms <- as.formula(paste0("survival::Surv(", var.time, ",", var.event, ") ~ ", paste(vars.ind, collapse = "+")))
cmodel <- NULL
if (is.null(design.survey)) {
if (!is.null(id.cluster)) {
cmodel <- survival::coxph(forms, data = data, y = T, cluster = get(id.cluster))
} else {
cmodel <- survival::coxph(forms, data = data, y = T)
}
} else {
cmodel <- survey::svycoxph(forms, design = design.survey, y = T)
}
lp <- stats::predict(cmodel, type = "lp")
vec.y <- sapply(cmodel$y, `[[`, 1)
out <- timeROC::timeROC(
T = vec.y[1:(length(vec.y) / 2)],
delta = vec.y[(length(vec.y) / 2 + 1):length(vec.y)],
marker = lp,
cause = 1,
weighting = "marginal",
times = t
)
if (out$AUC[2] < 0.5) {
out <- timeROC::timeROC(
T = vec.y[1:(length(vec.y) / 2)],
delta = vec.y[(length(vec.y) / 2 + 1):length(vec.y)],
marker = -lp,
cause = 1,
weighting = "marginal",
times = t
)
}
## Coxph object
data[[var.event]][data[[var.event]] == 1 & data[[var.time]] > t] <- 0
data[[var.time]][data[[var.time]] > t] <- t
if (is.null(design.survey)) {
if (!is.null(id.cluster)) {
cmodel <- survival::coxph(forms, data = data, y = T, cluster = get(id.cluster))
} else {
cmodel <- survival::coxph(forms, data = data, y = T)
}
} else {
cmodel <- survey::svycoxph(forms, design = design.survey, y = T)
}
return(list(coxph = cmodel, timeROC = out))
}
#' @title timeROC_table: extract AUC information from list of timeROChelper object.
#' @description extract AUC information from list of timeROChelper object.
#' @param ListModel list of timeROChelper object
#' @param dec.auc digits for AUC, Default: 3
#' @param dec.p digits for p value, Default: 3
#' @return table of AUC information
#' @details extract AUC information from list of timeROChelper object.
#' @examples
#' # library(survival)
#' # list.timeROC <- lapply(list("age", c("age", "sex")),
#' # function(x){
#' # timeROChelper("status", "time", x, t = 365, data = lung)
#' # })
#' # timeROC_table(list.timeROC)
#' @seealso
#' \code{\link[stats]{confint}}
#' \code{\link[data.table]{data.table}}
#' @rdname timeROC_table
#' @importFrom stats confint qnorm
#' @importFrom data.table data.table
#' @importFrom survival concordance
timeROC_table <- function(ListModel, dec.auc = 3, dec.p = 3) {
res.roc <- eval(parse(text = paste0("survival::concordance(", paste(paste0("lapply(ListModel, `[[`, 'coxph')[[", seq_along(ListModel), "]]"), collapse = ", "), ")")))
auc <- res.roc$concordance
se1.96 <- stats::qnorm(0.975) * sqrt(ifelse(length(ListModel) == 1, res.roc$var, diag(res.roc$var)))
auc.ci <- paste0(round(auc - se1.96, dec.auc), "-", round(auc + se1.96, dec.auc))
auc <- round(auc, dec.auc)
if (length(ListModel) == 1) {
out <- data.table::data.table(paste0("Model ", seq_along(ListModel)), auc, auc.ci)
names(out) <- c("Prediction Model", "AUC", "95% CI")
} else {
auc.pdiff <- c(NA, sapply(
seq_along(ListModel)[-1],
function(x) {
contr <- c(-1, 1)
dtest <- contr %*% res.roc$concordance[(x - 1):x]
dvar <- contr %*% res.roc$var[(x - 1):x, (x - 1):x] %*% contr
p <- 2 * pnorm(abs(dtest / sqrt(dvar)), lower.tail = F)
p <- ifelse(p < 0.001, "< 0.001", round(p, dec.p))
return(p)
}
))
out <- data.table::data.table(paste0("Model ", seq_along(ListModel)), auc, auc.ci, auc.pdiff)
names(out) <- c("Prediction Model", "AUC", "95% CI", "P-value for AUC Difference")
}
return(out[])
}
#' @title survIDINRI_helper: Helper function for IDI.INF.OUT in survIDINRI packages
#' @description Helper function for IDI.INF.OUT in survIDINRI packages
#' @param var.event event
#' @param var.time time
#' @param list.vars.ind list of independent variable
#' @param t time
#' @param data data
#' @param dec.auc digits for AUC, Default: 3
#' @param dec.p digits for p value, Default: 3
#' @param id.cluster cluster variable if marginal model, Default: NULL
#' @return IDI, NRI
#' @details Helper function for IDI.INF.OUT in survIDINRI packages
#' @examples
#' # library(survival)
#' # survIDINRI_helper("status", "time", list.vars.ind = list("age", c("age", "sex")),
#' # t = 365, data = lung)
#' @seealso
#' \code{\link[data.table]{data.table}}
#' \code{\link[stats]{model.matrix}}
#' \code{\link[survival]{coxph}}
#' \code{\link[survival]{Surv}}
#' \code{\link[survIDINRI]{IDI.INF.OUT}}
#' \code{\link[survIDINRI]{IDI.INF}}
#' @rdname survIDINRI_helper
#' @importFrom data.table data.table
#' @importFrom stats model.matrix
#' @importFrom survival coxph Surv
#' @importFrom survIDINRI IDI.INF.OUT IDI.INF
survIDINRI_helper <- function(var.event, var.time, list.vars.ind, t, data, dec.auc = 3, dec.p = 3, id.cluster = NULL) {
data <- data.table::data.table(data)
data[[var.event]] <- as.numeric(as.vector(data[[var.event]]))
vars <- c(Reduce(union, list(var.event, var.time, unlist(list.vars.ind))))
if (!is.null(id.cluster)) {
data <- na.omit(data[, .SD, .SDcols = c(vars, id.cluster)])
} else {
data <- na.omit(data[, .SD, .SDcols = vars])
}
mm <- lapply(
list.vars.ind,
function(x) {
if (!is.null(id.cluster)) {
stats::model.matrix(survival::coxph(as.formula(paste0("survival::Surv(", var.time, ",", var.event, ") ~ ", paste(x, collapse = "+"), "+ cluster(", id.cluster, ")")), data = data))
} else {
stats::model.matrix(survival::coxph(as.formula(paste0("survival::Surv(", var.time, ",", var.event, ") ~ ", paste(x, collapse = "+"))), data = data))
}
}
)
res <- lapply(
seq_along(list.vars.ind)[-1],
function(x) {
resIDINRI <- survIDINRI::IDI.INF.OUT(survIDINRI::IDI.INF(data[, .SD, .SDcols = c(var.time, var.event)], mm[[x - 1]], mm[[x]], t, npert = 200))
zz <- lapply(
list(resIDINRI[1, ], resIDINRI[2, ]),
function(x) {
c(round(x[1], dec.auc), paste0(round(x[2], dec.auc), "-", round(x[3], dec.auc)), ifelse(x[4] < 0.001, "< 0.001", round(x[4], dec.p)))
}
)
return(unlist(zz))
}
)
out <- data.table::data.table(Reduce(rbind, c(list(rep(NA, 6)), res)))
names(out) <- c("IDI", "95% CI", "P-value for IDI", "continuous NRI", "95% CI", "P-value for NRI")
return(out[])
}
#' @title timerocModule: shiny module server for time-dependent roc analysis
#' @description shiny module server for time-dependent roc analysis
#' @param input input
#' @param output output
#' @param session session
#' @param data Reactive data
#' @param data_label Reactuve data label
#' @param data_varStruct Reactive List of variable structure, Default: NULL
#' @param nfactor.limit nlevels limit in factor variable, Default: 10
#' @param design.survey Reactive survey data. default: NULL
#' @param id.cluster Reactive cluster variable if marginal model, Default: NULL
#' @param iid logical, get CI of AUC, Default: T
#' @param NRIIDI logical, get NRI & IDI, Default: T
#' @return shiny module server for time-dependent roc analysis
#' @details shiny module server for time-dependent roc analysis
#' @examples
#' library(shiny)
#' library(DT)
#' library(data.table)
#' library(jstable)
#' library(ggplot2)
#' library(timeROC)
#' library(survIDINRI)
#' ui <- fluidPage(
#' sidebarLayout(
#' sidebarPanel(
#' timerocUI("timeroc")
#' ),
#' mainPanel(
#' plotOutput("plot_timeroc"),
#' ggplotdownUI("timeroc"),
#' DTOutput("table_timeroc")
#' )
#' )
#' )
#'
#' server <- function(input, output, session) {
#' data <- reactive(mtcars)
#' data.label <- jstable::mk.lev(mtcars)
#'
#' out_timeroc <- callModule(timerocModule, "timeroc",
#' data = data, data_label = data.label,
#' data_varStruct = NULL
#' )
#'
#' output$plot_timeroc <- renderPlot({
#' print(out_timeroc()$plot)
#' })
#'
#' output$table_timeroc <- renderDT({
#' datatable(out_timeroc()$tb,
#' rownames = F, editable = F, extensions = "Buttons",
#' caption = "ROC results",
#' options = c(jstable::opt.tbreg("roctable"), list(scrollX = TRUE))
#' )
#' })
#' }
#' @seealso
#' \code{\link[stats]{quantile}}
#' \code{\link[data.table]{setkey}}
#' \code{\link[data.table]{data.table}}
#' \code{\link[data.table]{rbindlist}}
#' @rdname timerocModule
#' @export
#' @importFrom stats quantile median
#' @importFrom data.table setkey rbindlist data.table
#' @importFrom rvg dml
#' @importFrom officer read_pptx add_slide ph_with ph_location
#' @importFrom timeROC SeSpPPVNPV
timerocModule <- function(input, output, session, data, data_label, data_varStruct = NULL, nfactor.limit = 10, design.survey = NULL, id.cluster = NULL, iid = T, NRIIDI = T) {
## To remove NOTE.
ListModel <- compare <- level <- variable <- FP <- TP <- model <- Sensitivity <- Specificity <- NULL
if (is.null(data_varStruct)) {
data_varStruct <- reactive(list(variable = names(data())))
}
vlist <- reactive({
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
factor_vars <- names(data())[data()[, lapply(.SD, class) %in% c("factor", "character")]]
# factor_vars <- names(data())[sapply(names(data()), function(x){class(data()[[x]]) %in% c("factor", "character")})]
factor_list <- mklist(data_varStruct(), factor_vars)
conti_vars <- setdiff(names(data()), factor_vars)
if (!is.null(design.survey)) {
conti_vars <- setdiff(conti_vars, c(names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
}
conti_vars_positive <- conti_vars[unlist(data()[, lapply(.SD, function(x) {
min(x, na.rm = T) >= 0
}), .SDcols = conti_vars])]
conti_list <- mklist(data_varStruct(), conti_vars)
nclass_factor <- unlist(data()[, lapply(.SD, function(x) {
length(levels(x))
}), .SDcols = factor_vars])
# nclass_factor <- sapply(factor_vars, function(x){length(unique(data()[[x]]))})
class01_factor <- unlist(data()[, lapply(.SD, function(x) {
identical(levels(x), c("0", "1"))
}), .SDcols = factor_vars])
validate(
need(length(class01_factor) >= 1, "No categorical variables coded as 0, 1 in data")
)
factor_01vars <- factor_vars[class01_factor]
factor_01_list <- mklist(data_varStruct(), factor_01vars)
group_vars <- factor_vars[nclass_factor >= 2 & nclass_factor <= nfactor.limit & nclass_factor < nrow(data())]
group_list <- mklist(data_varStruct(), group_vars)
except_vars <- factor_vars[nclass_factor > nfactor.limit | nclass_factor == 1 | nclass_factor == nrow(data())]
return(list(
factor_vars = factor_vars, factor_list = factor_list, conti_vars = conti_vars, conti_list = conti_list, conti_vars_positive = conti_vars_positive,
factor_01vars = factor_01vars, factor_01_list = factor_01_list, group_vars = group_vars, group_list = group_list, except_vars = except_vars
))
})
output$eventtime <- renderUI({
validate(
need(length(vlist()$factor_01vars) >= 1, "No candidate event variables coded as 0, 1"),
need(length(vlist()$conti_vars_positive) >= 1, "No candidate time variables")
)
tagList(
selectInput(session$ns("event_km"), "Event",
choices = mklist(data_varStruct(), vlist()$factor_01vars), multiple = F,
selected = NULL
),
selectInput(session$ns("time_km"), "Time",
choices = mklist(data_varStruct(), vlist()$conti_vars_positive), multiple = F,
selected = NULL
)
)
})
nmodel <- reactiveVal(1)
output$addmodel <- renderUI({
if (nmodel() <= 1) {
actionButton(session$ns("add"), label = "Add model", icon("plus"), class = "btn-primary")
} else if (nmodel() > 1) {
tagList(
actionButton(session$ns("add"), label = "Add model", icon("plus"), class = "btn-primary"),
actionButton(session$ns("rmv"), label = "Remove model", icon("minus"))
)
}
})
indeproc <- reactive({
req(!is.null(input$event_km))
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
if (!is.null(design.survey)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
} else if (!is.null(id.cluster)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, id.cluster()))
} else {
indep.roc <- setdiff(names(data()), c(vlist()$except_vars, input$event_km))
}
return(indep.roc)
})
output$indep <- renderUI({
selectInput(session$ns(paste0("indep_km", 1)), paste0("Independent variables for Model ", 1),
choices = mklist(data_varStruct(), indeproc()), multiple = T,
selected = unlist(mklist(data_varStruct(), indeproc()))[1]
)
})
observeEvent(input$add, {
insertUI(
selector = paste0("div:has(> #", session$ns("add"), ")"),
where = "beforeBegin",
ui = selectInput(session$ns(paste0("indep_km", nmodel() + 1)), paste0("Independent variables for Model ", nmodel() + 1),
choices = mklist(data_varStruct(), indeproc()), multiple = T,
selected = unlist(mklist(data_varStruct(), indeproc()))[1:min(length(indeproc()), nmodel() + 1)]
)
)
nmodel(nmodel() + 1)
})
observeEvent(input$rmv, {
removeUI(
selector = paste0("div:has(>> #", session$ns(paste0("indep_km", nmodel())), ")")
)
nmodel(nmodel() - 1)
})
indeps <- reactive(lapply(1:nmodel(), function(i) {
input[[paste0("indep_km", i)]]
}))
output$time <- renderUI({
req(input$time_km)
tvar <- data()[[input$time_km]]
if (min(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
} else if (max(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 365, step = 5)
} else if (max(tvar, na.rm = T) >= 12) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 12)
} else {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
}
})
observeEvent(input$subcheck, {
output$subvar <- renderUI({
req(input$subcheck == T)
indeps.unique <- unique(unlist(indeps()))
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique))
if (!is.null(id.cluster)) {
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique, id.cluster()))
} else if (!is.null(design.survey)) {
var_subgroup <- setdiff(names(data()), union(c(names(design.survey()$strata), names(design.survey()$cluster), names(design.survey()$allprob)), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique)))
}
var_subgroup_list <- mklist(data_varStruct(), var_subgroup)
validate(
need(length(var_subgroup) > 0, "No variables for sub-group analysis")
)
tagList(
selectInput(session$ns("subvar_km"), "Sub-group variables",
choices = var_subgroup_list, multiple = T,
selected = var_subgroup[1]
)
)
})
})
output$subval <- renderUI({
req(input$subcheck == T)
req(length(input$subvar_km) > 0)
outUI <- tagList()
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
outUI[[v]] <- selectInput(session$ns(paste0("subval_km", v)), paste0("Sub-group value: ", input$subvar_km[[v]]),
choices = data_label()[variable == input$subvar_km[[v]], level], multiple = T,
selected = data_label()[variable == input$subvar_km[[v]], level][1]
)
} else {
val <- stats::quantile(data()[[input$subvar_km[[v]]]], na.rm = T)
outUI[[v]] <- sliderInput(session$ns(paste0("subval_km", v)), paste0("Sub-group range: ", input$subvar_km[[v]]),
min = val[1], max = val[5],
value = c(val[2], val[4])
)
}
}
outUI
})
timerocList <- reactive({
req(!is.null(input$event_km))
req(!is.null(input$time_km))
# req(!is.null(input$indep_km1))
# req(!is.null(input$indep_km2))
for (i in 1:nmodel()) {
req(!is.null(input[[paste0("indep_km", i)]]))
}
req(!is.null(indeps()))
collapse.indep <- sapply(1:nmodel(), function(i) {
paste0(input[[paste0("indep_km", i)]], collapse = "")
})
validate(
need(anyDuplicated(collapse.indep) == 0, "Please select different models")
)
data.km <- data()[complete.cases(data()[, .SD, .SDcols = unique(unlist(indeps()))])]
label.regress <- data_label()
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.km <- data.km[get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]]]
} else {
data.km <- data.km[get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2]]
}
}
data.km[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.km)[, c("variable", "level")]
data.table::setkey(data_label(), "variable", "level")
data.table::setkey(label.regress2, "variable", "level")
label.regress <- data_label()[label.regress2]
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
}
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km)
})
if ((nmodel() == 1 | NRIIDI == F)) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
res.cut <- NULL
}
} else {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km, id.cluster = id.cluster())
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km, id.cluster = id.cluster()
)
)
res.cut <- NULL
}
}
# res.tb <- timeROC_table(res.roc)
} else {
data.design <- design.survey()
label.regress <- data_label()
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.design <- subset(data.design, get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]])
} else {
data.design <- subset(data.design, get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2])
}
}
data.design$variables[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.design$variables)[, c("variable", "class", "level")]
data.table::setkey(data_label(), "variable", "class", "level")
data.table::setkey(label.regress2, "variable", "class", "level")
label.regress <- data_label()[label.regress2]
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
}
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km,
vars.ind = x,
t = input$time_to_roc, data = data.km, design.survey = data.design
)
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
res.cut <- NULL
}
}
res.timeROC <- lapply(res.roc, `[[`, "timeROC")
data.rocplot <- data.table::rbindlist(
lapply(
1:length(res.timeROC),
function(x) {
data.table::data.table(
FP = res.timeROC[[x]]$FP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
TP = res.timeROC[[x]]$TP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
model = paste0("model ", x)
)
}
)
)
p <- ggplot(data.rocplot, aes(FP, TP, colour = model)) +
geom_line() +
geom_abline(slope = 1, lty = 2) +
xlab("1-Specificity") +
ylab("Sensitivity")
return(list(plot = p, tb = res.tb, cut = res.cut))
})
output$downloadControls <- renderUI({
tagList(
column(
4,
selectizeInput(session$ns("file_ext"), "File extension (dpi = 300)",
choices = c("jpg", "pdf", "tiff", "svg", "pptx"), multiple = F,
selected = "pptx"
)
),
column(
4,
sliderInput(session$ns("fig_width"), "Width (in):",
min = 5, max = 15, value = 8
)
),
column(
4,
sliderInput(session$ns("fig_height"), "Height (in):",
min = 5, max = 15, value = 6
)
)
)
})
output$downloadButton <- downloadHandler(
filename = function() {
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
return(paste(input$event_km, "_", input$time_km, "_timeROC.", input$file_ext, sep = ""))
} else {
return(paste(input$event_km, "_", input$time_km, "_timeROC_marginal.", input$file_ext, sep = ""))
}
} else {
return(paste(input$event_km, "_", input$time_km, "__timeROC_survey.", input$file_ext, sep = ""))
}
},
# content is a function with argument file. content writes the plot to the device
content = function(file) {
withProgress(
message = "Download in progress",
detail = "This may take a while...",
value = 0,
{
for (i in 1:15) {
incProgress(1 / 15)
Sys.sleep(0.01)
}
if (input$file_ext == "pptx") {
my_vec_graph <- rvg::dml(ggobj = timerocList()$plot)
doc <- officer::read_pptx()
doc <- officer::add_slide(doc, layout = "Title and Content", master = "Office Theme")
doc <- officer::ph_with(doc, my_vec_graph, location = officer::ph_location(width = input$fig_width, height = input$fig_height))
print(doc, target = file)
} else {
ggsave(file, timerocList()$plot, dpi = 300, units = "in", width = input$fig_width, height = input$fig_height)
}
}
)
}
)
return(timerocList)
}
#' @title timerocModule2: shiny module server for time dependent roc analysis- input number of model as integer
#' @description shiny module server for time-dependent roc analysis- input number of model as integer
#' @param input input
#' @param output output
#' @param session session
#' @param data Reactive data
#' @param data_label Reactuve data label
#' @param data_varStruct Reactive List of variable structure, Default: NULL
#' @param nfactor.limit nlevels limit in factor variable, Default: 10
#' @param design.survey Reactive survey data. default: NULL
#' @param id.cluster Reactive cluster variable if marginal model, Default: NULL
#' @param iid logical, get CI of AUC, Default: T
#' @param NRIIDI logical, get NRI & IDI, Default: T
#' @return shiny module server for time dependent roc analysis- input number of model as integer
#' @details shiny module server for time dependent roc analysis- input number of model as integer
#' @examples
#' library(shiny)
#' library(DT)
#' library(data.table)
#' library(jstable)
#' library(ggplot2)
#' library(timeROC)
#' library(survIDINRI)
#' ui <- fluidPage(
#' sidebarLayout(
#' sidebarPanel(
#' timerocUI("timeroc")
#' ),
#' mainPanel(
#' plotOutput("plot_timeroc"),
#' ggplotdownUI("timeroc"),
#' DTOutput("table_timeroc")
#' )
#' )
#' )
#'
#' server <- function(input, output, session) {
#' data <- reactive(mtcars)
#' data.label <- jstable::mk.lev(mtcars)
#'
#' out_timeroc <- callModule(timerocModule2, "timeroc",
#' data = data, data_label = data.label,
#' data_varStruct = NULL
#' )
#'
#' output$plot_timeroc <- renderPlot({
#' print(out_timeroc()$plot)
#' })
#'
#' output$table_timeroc <- renderDT({
#' datatable(out_timeroc()$tb,
#' rownames = F, editable = F, extensions = "Buttons",
#' caption = "ROC results",
#' options = c(jstable::opt.tbreg("roctable"), list(scrollX = TRUE))
#' )
#' })
#' }
#' @seealso
#' \code{\link[stats]{quantile}}
#' \code{\link[data.table]{setkey}}
#' \code{\link[data.table]{data.table}}
#' \code{\link[data.table]{rbindlist}}
#' @rdname timerocModule
#' @export
#' @importFrom stats quantile median
#' @importFrom data.table setkey rbindlist data.table
#' @importFrom rvg dml
#' @importFrom officer read_pptx add_slide ph_with ph_location
timerocModule2 <- function(input, output, session, data, data_label, data_varStruct = NULL, nfactor.limit = 10, design.survey = NULL, id.cluster = NULL, iid = T, NRIIDI = T) {
## To remove NOTE.
ListModel <- compare <- level <- variable <- FP <- TP <- model <- Sensitivity <- Specificity <- NULL
if (is.null(data_varStruct)) {
data_varStruct <- reactive(list(variable = names(data())))
}
vlist <- reactive({
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
factor_vars <- names(data())[data()[, lapply(.SD, class) %in% c("factor", "character")]]
# factor_vars <- names(data())[sapply(names(data()), function(x){class(data()[[x]]) %in% c("factor", "character")})]
factor_list <- mklist(data_varStruct(), factor_vars)
conti_vars <- setdiff(names(data()), factor_vars)
if (!is.null(design.survey)) {
conti_vars <- setdiff(conti_vars, c(names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
}
conti_vars_positive <- conti_vars[unlist(data()[, lapply(.SD, function(x) {
min(x, na.rm = T) >= 0
}), .SDcols = conti_vars])]
conti_list <- mklist(data_varStruct(), conti_vars)
nclass_factor <- unlist(data()[, lapply(.SD, function(x) {
length(levels(x))
}), .SDcols = factor_vars])
# nclass_factor <- sapply(factor_vars, function(x){length(unique(data()[[x]]))})
class01_factor <- unlist(data()[, lapply(.SD, function(x) {
identical(levels(x), c("0", "1"))
}), .SDcols = factor_vars])
validate(
need(length(class01_factor) >= 1, "No categorical variables coded as 0, 1 in data")
)
factor_01vars <- factor_vars[class01_factor]
factor_01_list <- mklist(data_varStruct(), factor_01vars)
group_vars <- factor_vars[nclass_factor >= 2 & nclass_factor <= nfactor.limit & nclass_factor < nrow(data())]
group_list <- mklist(data_varStruct(), group_vars)
except_vars <- factor_vars[nclass_factor > nfactor.limit | nclass_factor == 1 | nclass_factor == nrow(data())]
return(list(
factor_vars = factor_vars, factor_list = factor_list, conti_vars = conti_vars, conti_list = conti_list, conti_vars_positive = conti_vars_positive,
factor_01vars = factor_01vars, factor_01_list = factor_01_list, group_vars = group_vars, group_list = group_list, except_vars = except_vars
))
})
output$eventtime <- renderUI({
validate(
need(length(vlist()$factor_01vars) >= 1, "No candidate event variables coded as 0, 1"),
need(length(vlist()$conti_vars_positive) >= 1, "No candidate time variables")
)
tagList(
selectInput(session$ns("event_km"), "Event",
choices = mklist(data_varStruct(), vlist()$factor_01vars), multiple = F,
selected = NULL
),
selectInput(session$ns("time_km"), "Time",
choices = mklist(data_varStruct(), vlist()$conti_vars_positive), multiple = F,
selected = NULL
)
)
})
output$addmodel <- renderUI({
radioButtons(session$ns("nmodel"), "Number of models", 1:5, selected = 1, inline = T)
})
nmodel <- reactive(as.integer(input$nmodel))
indeproc <- reactive({
req(!is.null(input$event_km))
mklist <- function(varlist, vars) {
lapply(
varlist,
function(x) {
inter <- intersect(x, vars)
if (length(inter) == 1) {
inter <- c(inter, "")
}
return(inter)
}
)
}
if (!is.null(design.survey)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, names(design.survey()$allprob), names(design.survey()$strata), names(design.survey()$cluster)))
} else if (!is.null(id.cluster)) {
indep.roc <- setdiff(vlist()$factor_vars, c(vlist()$except_vars, input$event_km, id.cluster()))
} else {
indep.roc <- setdiff(names(data()), c(vlist()$except_vars, input$event_km))
}
return(indep.roc)
})
output$indep <- renderUI({
req(nmodel())
lapply(1:nmodel(), function(x) {
selectInput(session$ns(paste0("indep_km", x)), paste0("Independent variables for Model ", x),
choices = mklist(data_varStruct(), indeproc()), multiple = T,
selected = unlist(mklist(data_varStruct(), indeproc()))[x]
)
})
})
indeps <- reactive(lapply(1:nmodel(), function(i) {
input[[paste0("indep_km", i)]]
}))
output$time <- renderUI({
req(input$time_km)
tvar <- data()[[input$time_km]]
if (min(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
} else if (max(tvar, na.rm = T) >= 365) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 365, step = 5)
} else if (max(tvar, na.rm = T) >= 12) {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = 12)
} else {
sliderInput(session$ns("time_to_roc"), "Time to analyze", min = min(tvar, na.rm = T), max = max(tvar, na.rm = T), value = median(tvar, na.rm = T))
}
})
observeEvent(input$subcheck, {
output$subvar <- renderUI({
req(input$subcheck == T)
indeps.unique <- unique(unlist(indeps()))
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique))
if (!is.null(id.cluster)) {
var_subgroup <- setdiff(names(data()), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique, id.cluster()))
} else if (!is.null(design.survey)) {
var_subgroup <- setdiff(names(data()), union(c(names(design.survey()$strata), names(design.survey()$cluster), names(design.survey()$allprob)), c(vlist()$except_vars, input$time_km, input$event_km, indeps.unique)))
}
var_subgroup_list <- mklist(data_varStruct(), var_subgroup)
validate(
need(length(var_subgroup) > 0, "No variables for sub-group analysis")
)
tagList(
selectInput(session$ns("subvar_km"), "Sub-group variables",
choices = var_subgroup_list, multiple = T,
selected = var_subgroup[1]
)
)
})
})
output$subval <- renderUI({
req(input$subcheck == T)
req(length(input$subvar_km) > 0)
outUI <- tagList()
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
outUI[[v]] <- selectInput(session$ns(paste0("subval_km", v)), paste0("Sub-group value: ", input$subvar_km[[v]]),
choices = data_label()[variable == input$subvar_km[[v]], level], multiple = T,
selected = data_label()[variable == input$subvar_km[[v]], level][1]
)
} else {
val <- stats::quantile(data()[[input$subvar_km[[v]]]], na.rm = T)
outUI[[v]] <- sliderInput(session$ns(paste0("subval_km", v)), paste0("Sub-group range: ", input$subvar_km[[v]]),
min = val[1], max = val[5],
value = c(val[2], val[4])
)
}
}
outUI
})
timerocList <- reactive({
req(!is.null(input$event_km))
req(!is.null(input$time_km))
# req(!is.null(input$indep_km1))
# req(!is.null(input$indep_km2))
for (i in 1:nmodel()) {
req(!is.null(input[[paste0("indep_km", i)]]))
}
req(!is.null(indeps()))
collapse.indep <- sapply(1:nmodel(), function(i) {
paste0(input[[paste0("indep_km", i)]], collapse = "")
})
validate(
need(anyDuplicated(collapse.indep) == 0, "Please select different models")
)
data.km <- data()
label.regress <- data_label()
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.km <- data.km[get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]]]
} else {
data.km <- data.km[get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2]]
}
}
data.km[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.km)[, c("variable", "class", "level")]
data.table::setkey(data_label(), "variable", "class", "level")
data.table::setkey(label.regress2, "variable", "class", "level")
label.regress <- data_label()[label.regress2]
data.km[[input$event_km]] <- as.numeric(as.vector(data.km[[input$event_km]]))
}
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km)
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
}
} else {
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km, vars.ind = x, t = input$time_to_roc, data = data.km, id.cluster = id.cluster())
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
troc <- timeROC::timeROC(
T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1,
weighting = "marginal",
times = input$time_to_roc, iid = F
)
mk <- data.km[[indeps()[[1]][1]]]
if (troc$AUC[2] < 0.5) {
mk <- -mk
}
res.cut <- data.table::rbindlist(lapply(unique(mk), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = mk,
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
if (troc$AUC[2] < 0.5) {
res.cut[, cut := -cut]
}
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km, id.cluster = id.cluster()
)
)
res.cut <- NULL
}
}
# res.tb <- timeROC_table(res.roc)
} else {
data.design <- design.survey()
label.regress <- data_label()
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
if (input$subcheck == TRUE) {
validate(
need(length(input$subvar_km) > 0, "No variables for subsetting"),
need(all(sapply(1:length(input$subvar_km), function(x) {
length(input[[paste0("subval_km", x)]])
})), "No value for subsetting")
)
for (v in seq_along(input$subvar_km)) {
if (input$subvar_km[[v]] %in% vlist()$factor_vars) {
data.design <- subset(data.design, get(input$subvar_km[[v]]) %in% input[[paste0("subval_km", v)]])
} else {
data.design <- subset(data.design, get(input$subvar_km[[v]]) >= input[[paste0("subval_km", v)]][1] & get(input$subvar_km[[v]]) <= input[[paste0("subval_km", v)]][2])
}
}
data.design$variables[, (vlist()$factor_vars) := lapply(.SD, factor), .SDcols = vlist()$factor_vars]
label.regress2 <- mk.lev(data.design$variables)[, c("variable", "class", "level")]
data.table::setkey(data_label(), "variable", "class", "level")
data.table::setkey(label.regress2, "variable", "class", "level")
label.regress <- data_label()[label.regress2]
data.design$variables[[input$event_km]] <- as.numeric(as.vector(data.design$variables[[input$event_km]]))
}
res.roc <- lapply(indeps(), function(x) {
timeROChelper(input$event_km, input$time_km,
vars.ind = x,
t = input$time_to_roc, data = data.km, design.survey = data.design
)
})
if (nmodel() == 1 | NRIIDI == F) {
res.tb <- timeROC_table(res.roc)
res.cut <- NULL
if (length(indeps()[[1]]) == 1) {
res.cut <- data.table::rbindlist(lapply(unique(data.km[[indeps()[[1]][1]]]), function(cut) {
zz <- timeROC::SeSpPPVNPV(
cutpoint = cut, T = data.km[[input$time_km]],
delta = data.km[[input$event_km]],
marker = data.km[[indeps()[[1]][1]]],
cause = 1, weighting = "marginal",
times = input$time_to_roc,
iid = F
)
return(data.table::data.table(cut = cut, Sensitivity = zz$TP[[2]], Specificity = 1 - zz$FP[[2]]))
}))[Sensitivity + Specificity == max(Sensitivity + Specificity)][1, ]
}
} else {
res.tb <- cbind(
timeROC_table(res.roc),
survIDINRI_helper(input$event_km, input$time_km,
list.vars.ind = indeps(),
t = input$time_to_roc,
data = data.km
)
)
res.cut <- NULL
}
}
res.timeROC <- lapply(res.roc, `[[`, "timeROC")
data.rocplot <- data.table::rbindlist(
lapply(
1:length(res.timeROC),
function(x) {
data.table::data.table(
FP = res.timeROC[[x]]$FP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
TP = res.timeROC[[x]]$TP[, which(res.timeROC[[x]]$times == input$time_to_roc)],
model = paste0("model ", x)
)
}
)
)
p <- ggplot(data.rocplot, aes(FP, TP, colour = model)) +
geom_line() +
geom_abline(slope = 1, lty = 2) +
xlab("1-Specificity") +
ylab("Sensitivity")
return(list(plot = p, tb = res.tb))
})
output$downloadControls <- renderUI({
tagList(
column(
4,
selectizeInput(session$ns("file_ext"), "File extension (dpi = 300)",
choices = c("jpg", "pdf", "tiff", "svg", "pptx"), multiple = F,
selected = "pptx"
)
),
column(
4,
sliderInput(session$ns("fig_width"), "Width (in):",
min = 5, max = 15, value = 8
)
),
column(
4,
sliderInput(session$ns("fig_height"), "Height (in):",
min = 5, max = 15, value = 6
)
)
)
})
output$downloadButton <- downloadHandler(
filename = function() {
if (is.null(design.survey)) {
if (is.null(id.cluster)) {
return(paste(input$event_km, "_", input$time_km, "_timeROC.", input$file_ext, sep = ""))
} else {
return(paste(input$event_km, "_", input$time_km, "_timeROC_marginal.", input$file_ext, sep = ""))
}
} else {
return(paste(input$event_km, "_", input$time_km, "__timeROC_survey.", input$file_ext, sep = ""))
}
},
# content is a function with argument file. content writes the plot to the device
content = function(file) {
withProgress(
message = "Download in progress",
detail = "This may take a while...",
value = 0,
{
for (i in 1:15) {
incProgress(1 / 15)
Sys.sleep(0.01)
}
if (input$file_ext == "pptx") {
my_vec_graph <- rvg::dml(ggobj = timerocList()$plot)
doc <- officer::read_pptx()
doc <- officer::add_slide(doc, layout = "Title and Content", master = "Office Theme")
doc <- officer::ph_with(doc, my_vec_graph, location = officer::ph_location(width = input$fig_width, height = input$fig_height))
print(doc, target = file)
} else {
ggsave(file, timerocList()$plot, dpi = 300, units = "in", width = input$fig_width, height = input$fig_height)
}
}
)
}
)
return(timerocList)
}
Try the jsmodule package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.