Nothing
R/helper.r
In Rcan: Cancer Registry Data Analysis and Visualisation
Defines functions
core.csu_group_cases
core.csu_icd_ungroup
core.csu_icd_group
core.csu_year_extract
core.csu_ageSpecific_top
core.csu_ageSpecific
core.csu_eapc
core.csu_cumrisk
core.csu_asr
core.csu_format_export
core.csu_year_tick_generator
core.csu_axes_label
core.csu_tick_generator
core.csu_legend_wrapper
core.csu_dt_rank
core.error_time_variable
core.error_age_parse
core.error_missingage
core.error_variable
core.error_variable <- function(df_data, varname, funcname,type="numeric") {
if (!is.null(varname)) {
if (!(varname%in% colnames(df_data))) {
stop(paste0(varname, " is not a column of ", deparse(substitute(df_data)), " ,see documentation: Help(", deparse(substitute(funcname)), ")"))
}
if (type == "numeric") {
if (!is.numeric(df_data[[varname]])) {
stop(paste0(varname, " type must be ",type," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
} else if (type == "character"){
if (!is.character(df_data[[varname]])) {
stop(paste0(varname, " type must be ",type," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
} else if (type == "factor"){
if (!is.factor(df_data[[varname]])) {
stop(paste0(varname, " type must be ",type," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
}
}
}
core.error_missingage <- function(df_data, var_age,missing_age,funcname) {
if (!missing_age %in% df_data[[var_age]]) {
stop(paste0("\n",missing_age, " value is not present in the variable ",var_age," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
}
core.error_age_parse <- function(df_data, var_age, missing_age,funcname) {
dt_data <- as.data.table(df_data)
dt_data <- dt_data[, c(var_age), with=FALSE]
setnames(dt_data, var_age, 'CSU_A')
if (!is.null(missing_age)) {
dt_data[CSU_A == missing_age, CSU_A := NA]
}
if (!all(grepl(".*?(\\d{1,3}).*$",dt_data[!is.na(CSU_A)]$CSU_A, perl=TRUE)))
{
temp <- unique(as.character(dt_data[!grepl(".*?(\\d{1,3}).*$",dt_data[!is.na(CSU_A)]$CSU_A, perl=TRUE),]$CSU_A))
stop(paste0("\nthe value ",temp,", in the variable ",var_age, "\ncannot be change to numeric and is not declared as missing age,\nSee documentation: Help(", deparse(substitute(funcname)), ")\n"))
}
}
core.error_time_variable <- function(df_data, var_year, group_by, funcname) {
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
}
dt_data <- data.table(df_data, key = c(group_by))
dt_data$temp <- 1
nrow_base <- nrow(dt_data)
dt_test <- dt_data[ ,temp:=sum(temp), by=c(var_year, group_by)]
nrow_test <- nrow(dt_data[ ,sum(temp), by=c(var_year, group_by)])
if (nrow_test != nrow_base) {
setkeyv(dt_test, c(group_by,var_year))
print(head(dt_test[temp>1, ]))
dt_data <- NULL
stop(paste0("There is more than 1 data per year (see above).\nUse the 'group_by' option"," ,see documentation: Help(", deparse(substitute(funcname)), ")\n","or call the function on a subset to define the sub-population of interest.\n"))
}
}
core.csu_dt_rank <- function(df_data,
var_value = "CASES",
var_rank = "cancer_label",
group_by = NULL,
number = NULL,
ties.method="min") {
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
df_data <- as.data.table(df_data)
dt_rank <- df_data[, list(rank_value=sum(get(var_value))), by=c(var_rank, group_by)]
dt_rank[, CSU_RANK:= frank(-rank_value, ties.method=ties.method), by=group_by]
if (!is.null(number)){
dt_rank <- dt_rank[CSU_RANK <= number,c(group_by, var_rank, "CSU_RANK"), with=FALSE]
}
df_data <- merge(dt_rank, df_data,by=c(group_by, var_rank), all.x=TRUE)
if (bool_dum_by) {
df_data[,CSU_dum_by:=NULL]
}
return(df_data)
}
core.csu_legend_wrapper <- function(label, width) {
label <- sapply(strwrap(label, width = width, simplify = FALSE), paste, collapse="\n")
return(label)
}
core.csu_tick_generator <- function(max,min = 0,logscale=FALSE) {
if (!logscale) {
if (min > 0) {
min = 0
}
if (max < 0) {
max = 0
}
temp_log_max = 10^floor(log10(max-min))
temp_unit_floor_max = floor((max-min)/(temp_log_max))
if (temp_unit_floor_max < 2) {
tick_space = 0.2*temp_log_max
} else {
if (temp_unit_floor_max < 5) {
tick_space = 0.5*temp_log_max
} else {
tick_space = temp_log_max
}
}
temp_top <- ceiling(max/tick_space)*tick_space
temp_floor <- floor(min/tick_space)*tick_space
tick_list <- seq(temp_floor, temp_top, tick_space)
tick_minor_list <- NULL
} else {
temp_log_max = 10^floor(log10(max))
temp_unit_floor_max = floor(max/(temp_log_max))
temp_log_min <- 10^floor(log10(min))
temp_unit_floor_min <- floor(min/(temp_log_min))
if (temp_log_min == temp_log_max) {
tick_list <- c(temp_unit_floor_min:(temp_unit_floor_max+1)*temp_log_min)
if (temp_unit_floor_max == temp_unit_floor_min) {
tick_minor_list <- c((temp_unit_floor_min*temp_log_min)+0:9*(temp_log_min/10))
} else {
tick_minor_list <- c((temp_unit_floor_min*temp_log_min)+0:19*(temp_log_min/10))
}
} else if (temp_log_max/temp_log_min < 1000) {
if (temp_unit_floor_min < 6) {
tick_list <- temp_unit_floor_min:5*temp_log_min
tick_list <- c(tick_list, temp_log_min*7) ## min . . 5 7
} else {
tick_list <- temp_unit_floor_min*temp_log_min ## min
}
tick_minor_list <- temp_unit_floor_min:19*temp_log_min ## min . . 19
while (temp_log_min != (temp_log_max/10)) {
temp_log_min = temp_log_min*10
tick_list <- c(tick_list, c(1,2,3,5,7)*temp_log_min)
tick_minor_list <- c(tick_minor_list, 2:19*temp_log_min)
}
tick_minor_list <- c(tick_minor_list, 2:(temp_unit_floor_max+1)*temp_log_max)
if (temp_unit_floor_max <5) {
tick_list <- c(tick_list, 1:(temp_unit_floor_max+1)*temp_log_max)
} else if (temp_unit_floor_max <7) {
tick_list <- c(tick_list, c(1,2,3,5,temp_unit_floor_max+1)*temp_log_max)
} else {
tick_list <- c(tick_list, c(1,2,3,5,7,temp_unit_floor_max+1)*temp_log_max)
}
} else {
if (temp_unit_floor_min == 1) {
tick_list <- c(1,2,3,5)*temp_log_min
} else if (temp_unit_floor_min == 2) {
tick_list <- c(2,3,5)*temp_log_min
} else if (temp_unit_floor_min < 6) {
tick_list <- c(5,7)*temp_log_min
} else {
tick_list <- 7*temp_log_min
}
tick_minor_list <- temp_unit_floor_min:9*temp_log_min ## min . . 19
while (temp_log_min != (temp_log_max/10)) {
temp_log_min = temp_log_min*10
tick_list <- c(tick_list, c(1,2,5)*temp_log_min)
tick_minor_list <- c(tick_minor_list, 2:9*temp_log_min)
}
tick_minor_list <- c(tick_minor_list, 2:(temp_unit_floor_max+1)*temp_log_max)
if (temp_unit_floor_max <5) {
tick_list <- c(tick_list, unique(c(1,2,temp_unit_floor_max+1)*temp_log_max))
} else if (temp_unit_floor_max <6) {
tick_list <-c(tick_list, c(1,2,5)*temp_log_max)
} else if (temp_unit_floor_max < 7) {
tick_list <- c(tick_list, c(1,2,5,7)*temp_log_max)
} else {
tick_list <- c(tick_list, c(1,2,5,7,temp_unit_floor_max+1)*temp_log_max)
}
}
}
return(list(tick_list=tick_list, tick_minor_list=tick_minor_list))
}
core.csu_axes_label <- function(l) {
l <- format(l, big.mark = ",", scientific = FALSE, drop0trailing = TRUE)
}
core.csu_year_tick_generator <- function(min, max) {
mod <- 5
if (max - min < 10 ) {
mod <- 1
} else if (max - min < 20){
mod <- 2
}
temp1 <- min - (min %% mod)
temp2 <- max - (max %% mod) +ifelse(mod>=5,mod,0)
if (temp2 - temp1 <= mod*6) {
year_space <- mod
year_list <- seq(temp1,temp2,year_space)
year_minor_list <- year_list
} else {
year_space <- mod*2
if (temp1 %% mod*2 > 0) {
year_list <- seq(temp1+mod,temp2,year_space)
year_minor_list <- seq(temp1,temp2,year_space/2)
} else {
year_list <- seq(temp1,temp2,year_space)
year_minor_list <- seq(temp1,temp2,year_space/2)
}
}
return(list(tick_list=year_list, tick_minor_list=year_minor_list))
}
core.csu_format_export <- function(type, plot_title, landscape=FALSE) {
#http://www.altelia.fr/actualites/calculateur-resolution-definition-format.htm
# 6 inch = 15.24 cm
#10,795
#format
#if (!is.null(format_export)) {
# filename <- gsub("[[:punct:][:space:]\n]", "_", plot_title)
# core.csu_format_export(format_export, plot_title = filename, landscape = FALSE)
#}
# help file
# \item{format_export}{
# export the graph in different format:\cr
# \tabular{ll}{
# \code{NULL} \tab Plot in R studio windows.\cr
# \code{"pdf"} \tab Export in PDF format.\cr
# \code{"tiff"} \tab Export in TIFF 300dpi format.\cr
# \code{"png"} \tab Export in PNG 200dpi format.\cr
# \code{"svg"} \tab Export in SVG format. Can be edit with \url{https://inkscape.org/fr/}.\cr
# }
# The filename is the \code{plot_title} option.
# }
# \item{graph_dev}{
# If the plot is embedded in a graphics Device function (such as pdf()), the graph_dev option should be set to TRUE for the first graph to avoid a blank page.
# }
png_width <- ifelse(landscape, 2339 , 1654 )
png_height <- ifelse(landscape, 1654 , 2339 )
tiff_width <- ifelse(landscape, 3508 , 2480 )
tiff_height <- ifelse(landscape, 2480 , 3508 )
pdf_width <- ifelse(landscape, 11.692 , 8.267 )
pdf_height <- ifelse(landscape, 8.267 , 11.692 )
if (type == "pdf") {
pdf(paste(plot_title,".",type, sep=""),width = pdf_width, height = pdf_height)
} else if (type == "svg") {
svg(paste(plot_title,".",type, sep=""),width = pdf_width, height = pdf_height)
} else if (type == "png") {
png(paste(plot_title,".",type, sep=""),width = png_width, height = png_height, units = "px",res = 200)
} else if (type == "tiff") {
tiff(paste(plot_title,".",type, sep=""),width = tiff_width, height = tiff_height,units = "px",res = 300,compression ="lzw")
}
}
core.csu_asr <- function(df_data, var_age, var_cases, var_py, group_by=NULL,
var_age_group=NULL, missing_age = NULL,db_rate = 100000,
first_age = 1, last_age = 18, pop_base = "SEGI", crude_rate = FALSE,
var_st_err=NULL,correction_info=FALSE, var_asr="asr",age_dropped = FALSE,
pop_base_count = NULL, age_label_list = NULL, Rcan_print=FALSE)
{
bool_dum_by <- FALSE
bool_dum_age <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
if (is.null(var_age_group)) {
df_data$CSU_dum_age <- "dummy_age_gr"
var_age_group <- "CSU_dum_age"
group_by <- c(group_by, "CSU_dum_age")
bool_dum_age <- TRUE
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_age, "CSU_A")
setnames(dt_data, var_cases, "CSU_C")
setnames(dt_data, var_py, "CSU_P")
if (crude_rate == FALSE)
{
if (is.null(pop_base_count)) {
dt_data <- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P)), by=c(group_by, "CSU_A")]
}
# create index to keep order
index_order <- c(1:nrow(dt_data))
dt_data$index_order <- index_order
# missing age
dt_data[dt_data$CSU_A %in% missing_age,CSU_A:=NA ]
dt_data[is.na(dt_data$CSU_A),CSU_P:=0 ]
#parse age
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
#create age dummy: 1 2 3 4 --- 19
dt_data$age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
# correction factor
dt_data$correction <- 1
if (!is.null(missing_age)) {
dt_data[, total:=sum(CSU_C), by=group_by] #add total
dt_data[!is.na(dt_data$age_factor) , total_known:=sum(CSU_C), by=group_by] #add total_know
dt_data$correction <- dt_data$total / dt_data$total_know
dt_data[is.na(dt_data$correction),correction:=1 ]
dt_data$total <- NULL
dt_data$total_known <- NULL
}
if (is.null(pop_base_count)) {
# create world population DF for different nb of age group
SEGI_pop <- c(12000,10000,9000,9000,8000,8000,6000,6000,6000,6000,5000,4000,4000,3000,2000,1000,500,500)
EURO_pop <- c(8000,7000,7000,7000,7000,7000,7000,7000,7000,7000,7000,6000,5000,4000,3000,2000,1000,1000)
EURO2_pop <- c(5000,5500,5500,5500,6000,6000,6500,7000,7000,7000,7000,6500,6000,5500,5000,4000,2500,2500)
WHO_pop <- c(8860,8690,8600,8470,8220,7930,7610,7150,6590,6040,5370,4550,3720,2960,2210,1520,910,630)
if (pop_base == "EURO") {
pop <- EURO_pop
} else if (pop_base == "EURO2") {
pop <- EURO2_pop
} else if (pop_base == "WHO") {
pop <- WHO_pop
} else {
pop <- SEGI_pop
}
# calculated total pop for age selected
total_pop <- sum(pop[first_age:last_age])
Standard_pop <- data.table(pop = pop, age_factor= c(1:18))
pop[17] <- pop[17]+ pop[18]
pop[18] <- 0
Standard_pop$pop17 <- pop
pop[16] <- pop[16]+ pop[17]
pop[17] <- 0
Standard_pop$pop16 <- pop
pop[15] <- pop[15]+ pop[16]
pop[16] <- 0
Standard_pop$pop15 <- pop
#age dropped option
if (age_dropped) {
dt_data$age_factor <- dt_data$age_factor + first_age -1
}
# keep age selected
dt_data=dt_data[dt_data$age_factor %in% c(first_age:last_age) | is.na(dt_data$age_factor), ]
# calculated maximum age group with population data
if (last_age == 18) {
dt_data <- merge(dt_data, dt_data[dt_data$CSU_P != 0,list(nb_age_group = max(age_factor)), by=var_age_group], by=var_age_group)
} else {
dt_data$nb_age_group <- 18
}
# show population with less than 18 age group
if (last_age == 18) {
temp <- subset(dt_data,nb_age_group <18, select= c(var_age_group, "nb_age_group"))
if (nrow(temp) >0) {
setkey(temp,NULL)
if (Rcan_print) {
cat("\n")
cat("Population with less than 18 age group:\n\n" )
print
print(unique(temp), row.names = FALSE)
cat("\n")
}
}
temp <- NULL
}
#regroup case for population with nb of age group < 18
for (i in 15:17) {
if (i %in% dt_data$nb_age_group) {
dt_data[nb_age_group == i & age_factor >= i , CSU_C:=sum(CSU_C), by=group_by] #add total_know
dt_data[nb_age_group == i & age_factor > i & !is.na(age_factor), CSU_C := 0]
}
}
#add world pop to database
dt_data <- merge(dt_data,Standard_pop, by =c("age_factor"), all.x=TRUE )
Standard_pop <- NULL
dt_data[nb_age_group==17, pop:=pop17]
dt_data[nb_age_group==16, pop:=pop16]
dt_data[nb_age_group==15, pop:=pop15]
#return(dt_data)
} else {
#keep age group selected
dt_data <- dt_data[age_factor %in% (first_age:last_age), ]
#calcul total pop for canreg
total_pop <-sum(unique(dt_data[, c("age_factor", pop_base_count), with=FALSE])[[pop_base_count]])
#get age group list variable
if (is.null(age_label_list)) {
age_label_list <- var_age
}
age_group_list <- as.character(unique(dt_data[[age_label_list]]))
age_group_list <- paste(age_group_list, collapse=" ")
#rename variable population reference
setnames(dt_data, pop_base_count, "pop")
}
#calcul ASR
dt_data[dt_data$CSU_P != 0,rate:= dt_data$CSU_C[dt_data$CSU_P != 0]/ dt_data$CSU_P[dt_data$CSU_P != 0] * db_rate]
dt_data$asr <- dt_data$rate * dt_data$pop
dt_data[is.na(dt_data$asr),asr:=0 ]
dt_data$st_err <- ( dt_data$rate * (dt_data$pop^2) * (db_rate - dt_data$rate))/dt_data$CSU_P
dt_data[is.na(dt_data$st_err),st_err:=0 ]
# to check order
dt_data<- dt_data[order(dt_data$index_order ),]
dt_data<- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P),asr=sum(asr),st_err = sum(st_err),correction = max(correction)), by=group_by]
dt_data$asr <- dt_data$asr / total_pop
dt_data$asr <- dt_data$asr * dt_data$correction
dt_data$st_err <- (dt_data$st_err / (total_pop^2))^(1/2)
dt_data$st_err <- dt_data$st_err * dt_data$correction
dt_data$asr <- round(dt_data$asr, digits = 2)
dt_data$st_err <- round(dt_data$st_err, digits = 2)
dt_data$correction <- round((dt_data$correction-1)*100, digits = 1)
if (is.null(var_st_err)) {
dt_data$st_err <- NULL
} else {
setnames(dt_data, "st_err", var_st_err)
}
if (var_asr!="asr") {
setnames(dt_data, "asr", var_asr)
}
if (!correction_info) {
dt_data$correction <- NULL
}
}
else
{
#parse age
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
#create age dummy: 1 2 3 4 --- 19
dt_data$age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
#age dropped option
if (age_dropped) {
dt_data$age_factor <- dt_data$age_factor + first_age -1
}
# keep age selected
dt_data=dt_data[dt_data$age_factor %in% c(first_age:last_age) | is.na(dt_data$age_factor), ]
dt_data <- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P)), by=c(group_by)]
dt_data <- dt_data[, CSU_RESULT := CSU_C/CSU_P*db_rate]
if (var_asr=="asr") {
setnames(dt_data, "CSU_RESULT", 'crude_rate')
}
else
{
setnames(dt_data, "CSU_RESULT", var_asr)
}
}
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
if (bool_dum_age) {
df_data$CSU_dum_age <- NULL
}
if (bool_dum_by) {
df_data$CSU_dum_by <- NULL
}
setnames(df_data, "CSU_C", var_cases)
setnames(df_data, "CSU_P", var_py)
if (is.null(pop_base_count)) {
if (Rcan_print) {
temp <- last_age*5-1
if (last_age == 18) temp <- "99+"
if (crude_rate == FALSE)
{
cat("ASR have been computed for the age group ", (first_age-1)*5,"-", temp , "\n", sep="" )
}
else {
cat("Crude rate have been computed for the age group ", (first_age-1)*5,"-", temp , "\n", sep="" )
}
}
temp<- NULL
} else {
#cat("ASR have been computed for the age groups:\n",age_group_list , "\n", sep="" )
age_group_list<- NULL
}
return(df_data)
}
core.csu_cumrisk <- function(df_data, var_age, var_cases, var_py, group_by=NULL,
missing_age = NULL, last_age = 18,var_st_err=NULL,correction_info=FALSE,
var_cumrisk="cumrisk", age_label_list = NULL, Rcan_print=FALSE)
{
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_age, "CSU_A")
setnames(dt_data, var_cases, "CSU_C")
setnames(dt_data, var_py, "CSU_P")
dt_data <- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P)), by=c(group_by, "CSU_A")]
# create index to keep order
index_order <- c(1:nrow(dt_data))
dt_data$index_order <- index_order
# missing age
dt_data[dt_data$CSU_A %in% missing_age,CSU_A:=NA ]
dt_data[is.na(dt_data$CSU_A),CSU_P:=0 ]
#parse age
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
#create age dummy: 1 2 3 4 --- 19
dt_data$age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
# correction factor
dt_data$correction <- 1
if (!is.null(missing_age)) {
dt_data[, total:=sum(CSU_C), by=group_by] #add total
dt_data[!is.na(dt_data$age_factor) , total_known:=sum(CSU_C), by=group_by] #add total_know
dt_data$correction <- dt_data$total / dt_data$total_know
dt_data[is.na(dt_data$correction),correction:=1 ]
dt_data$total <- NULL
dt_data$total_known <- NULL
dt_data<- dt_data[!is.na(age_factor),]
}
if (!is.null(age_label_list)) {
# calcul year interval from age group label
dt_temp <- unique(dt_data[, c(age_label_list), with=FALSE])
dt_temp[, min:=as.numeric(regmatches(get(age_label_list), regexpr("[0-9]+",get(age_label_list))))]
dt_temp[, max:=shift(min, type ="lead")]
dt_temp[, age_span := max-min]
dt_temp <- dt_temp[, c("age_span",age_label_list), with=FALSE]
dt_data <- merge(dt_data, dt_temp,by= age_label_list, all.x=TRUE)
} else {
dt_data[, age_span:=5]
}
#keep age group selected
age_max <- max(dt_data$age_factor)
if (age_max-1 < last_age) {
last_age <- age_max-1
}
dt_data=dt_data[dt_data$age_factor <= last_age]
#calculate cum risk
dt_data[,cumrisk:=age_span*(CSU_C/CSU_P)]
dt_data[CSU_P==0,cumrisk:=0]
dt_data[, st_err := (CSU_C/CSU_P)/CSU_P]
dt_data[is.na(dt_data$st_err),st_err:=0 ]
# to check order
dt_data<- dt_data[order(dt_data$index_order ),]
dt_data <- dt_data[,list( cumrisk=sum(cumrisk), CSU_P=sum(CSU_P),CSU_C=sum(CSU_C),st_err = sum(st_err),correction = max(correction)), by=group_by]
dt_data[,cumrisk:=(1-exp(-cumrisk))*100*correction]
dt_data[, st_err:=(st_err^(1/2))*100*5]
dt_data[, st_err:=st_err*correction]
dt_data[,cumrisk:=round(cumrisk, digits = 2)]
dt_data[,st_err:=round(st_err, digits = 2)]
dt_data[, correction:=round((correction-1)*100, digits = 1)]
if (is.null(var_st_err)) {
dt_data$st_err <- NULL
} else {
setnames(dt_data, "st_err", var_st_err)
}
if (var_cumrisk!="cumrisk") {
setnames(dt_data, "cumrisk", var_cum_risk)
}
if (!correction_info) {
dt_data$correction <- NULL
}
setnames(dt_data, "CSU_C", var_cases)
setnames(dt_data, "CSU_P", var_py)
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
if (bool_dum_by) {
df_data$CSU_dum_by <- NULL
}
if (Rcan_print) {
temp <- last_age*5-1
cat("Cumulative risk have been computed for the age group ","0-", temp , "\n", sep="" )
temp<- NULL
}
return(df_data)
}
core.csu_eapc <- function(df_data,
var_rate="asr",
var_year="year",
group_by= NULL,
var_eapc="eapc",
CI_level = 0.95)
{
#create fake group to have group_by optional
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
dt_data <- data.table(df_data, key = c(group_by))
setnames(dt_data, var_rate, "CSU_R")
setnames(dt_data, var_year, "CSU_Y")
# remove 0 data
dt_data <- dt_data[!CSU_R == 0,]
dt_data[, id_group:=.GRP, by=group_by]
temp_max <- max(dt_data$id_group)
for (i in 1:temp_max) {
temp <- summary(glm(log(CSU_R) ~ CSU_Y,
family=gaussian(link = "identity"),
data=dt_data[dt_data$id_group == i,]
))
dt_data[dt_data$id_group == i, CSU_EAPC:=temp$coefficients[[2]]]
dt_data[dt_data$id_group == i, CSU_ST:=temp$coefficients[[4]]]
}
dt_data$CSU_UP <- 100*(exp(dt_data$CSU_EAPC-(qt((1-CI_level)/2, df = Inf)*dt_data$CSU_ST))-1)
dt_data$CSU_LOW <- 100*(exp(dt_data$CSU_EAPC+(qt((1-CI_level)/2, df = Inf)*dt_data$CSU_ST))-1)
dt_data$CSU_EAPC <- 100*(exp(dt_data$CSU_EAPC)-1)
dt_data<- dt_data[,list( CSU_EAPC=mean(CSU_EAPC), CSU_UP=mean(CSU_UP),CSU_LOW=mean(CSU_LOW)), by=group_by]
setnames(dt_data, "CSU_EAPC", var_eapc)
setnames(dt_data, "CSU_UP", paste(var_eapc, "up", sep="_"))
setnames(dt_data, "CSU_LOW", paste(var_eapc, "low", sep="_"))
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
if (bool_dum_by) {
df_data$CSU_dum_by <- NULL
}
return(df_data)
}
core.csu_ageSpecific <-function(df_data,
var_age="age",
var_cases="cases",
var_py="py",
group_by = NULL,
missing_age = NULL,
db_rate = 100000,
logscale=FALSE,
plot_title=NULL,
legend=csu_trend_legend(),
color_trend = NULL,
CI5_comparison=NULL,
linesize = 0.5,
age_label_list = NULL,
log_point=TRUE,
plot_subtitle=NULL,
plot_caption=NULL,
xtitle = "Age at diagnosis",
ytitle = "Age-specific incidence rate per",
label_group_by = waiver())
{
bool_CI5_comp <- FALSE
CI5_cancer_label <- NULL
bool_dum_by <- FALSE
# manage CI5 import
if (!is.null(CI5_comparison)) {
bool_CI5_comp <- TRUE
data(csu_ci5_mean, envir = e <- new.env())
df_CI5 <- e$csu_ci5_mean
dt_CI5 <- data.table(df_CI5)
if (is.character(CI5_comparison)) {
if (!(CI5_comparison%in% dt_CI5$ci5_cancer_label)) {
stop('CI5_comparison value must be a correct cancer label, see documentation: Help(csu_ci5_mean)')
} else {
dt_CI5 <- dt_CI5[dt_CI5$ci5_cancer_label == CI5_comparison, ]
}
} else {
if (is.numeric(CI5_comparison)) {
if (!(CI5_comparison%in% dt_CI5$ci5_cancer_code)) {
stop('CI5_comparison value must be a correct cancer code, see documentation: Help(csu_ci5_mean)')
} else {
dt_CI5 <- dt_CI5[dt_CI5$ci5_cancer_code == CI5_comparison, ]
}
}
}
CI5_cancer_label <- toString(dt_CI5$ci5_cancer_label[1])
}
# manage group by options
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_age, "CSU_A")
setnames(dt_data, var_cases, "CSU_C")
setnames(dt_data, var_py, "CSU_P")
setnames(dt_data, group_by, "CSU_BY")
##group population (use sum)
dt_data <- dt_data[, list(CSU_C=sum(CSU_C),CSU_P=sum(CSU_P)), by=c("CSU_BY", "CSU_A") ]
##change by to factor
dt_data$CSU_BY <- factor(dt_data$CSU_BY)
##to calcul age group
dt_data[CSU_A %in% missing_age,CSU_A:=NA ]
dt_data[is.na(CSU_A),CSU_P:=0 ]
#parse age
dt_data[, temp_label:=CSU_A]
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
dt_temp_label <- unique(dt_data[, c("CSU_A", "temp_label"), with=FALSE])
dt_data[, temp_label:= NULL]
dt_data$CSU_age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
dt_data <- merge(dt_data, dt_data[dt_data$CSU_P != 0,list(nb_age_group = max(CSU_age_factor)), by="CSU_BY"], by="CSU_BY")
for (i in 15:17) {
if (i %in% dt_data$nb_age_group) {
dt_data[nb_age_group == i & CSU_age_factor >= i , CSU_C:=sum(CSU_C), by="CSU_BY"] ##add total_know
dt_data[nb_age_group == i & CSU_age_factor > i & !is.na(CSU_age_factor), CSU_C := 0]
}
}
dt_data <- dt_data[CSU_P!=0]
##calcul rate
dt_data$rate <- dt_data$CSU_C/dt_data$CSU_P *db_rate
##create age label:
if (is.null(age_label_list)) {
##create age dummy: 1 2 3 4 --- 18
##regroup case for population with nb of age group < 18
max_age <- max(dt_data$nb_age_group)
if (max_age > 18) {
stop('The function cannot have more than 18 age-group, see documentation: Help(csu_graph_ageSpecific)')
}
age_label <- c("0-4","5-9","10-14","15-19","20-24","25-39","30-34","35-39","40-44", "45-49","50-54","55-59","60-64","65-69","70-74","75-79","80-84","85+")
age_label <- c(age_label[1:(max_age-1)], paste0((max_age-1)*5,"+"))
} else {
age_label <-age_label_list
max_age <- length(age_label)
}
## to calcul breaks
if (logscale) {
min_tick_value <- min(dt_data[rate != 0,]$rate)
} else {
min_tick_value <- 0
}
tick <- core.csu_tick_generator(max = max(dt_data$rate), min=min_tick_value, logscale = logscale )
tick_space <- tick$tick_list[length(tick$tick_list)] - tick$tick_list[length(tick$tick_list)-1]
temp_top <- ceiling(max(dt_data$rate)/tick_space)*tick_space
temp_expand_y <- max(dt_data$rate)/35
temp_expand_y_up <- max(dt_data$rate)+temp_expand_y
if (temp_expand_y_up > temp_top-(tick_space/2)) {
temp_expand_y_up <- temp_top+temp_expand_y
}
th_legend <- list(theme(legend.position="none"))
if (!bool_dum_by & legend$position == "bottom") {
th_legend <- list(theme(
legend.key = element_rect(fill="transparent"),
legend.position = "bottom",
legend.text = element_text(size = 14),
legend.title = element_text(size = 14),
legend.key.size=unit(1,"cm"),
legend.margin = margin(0, 0, 0, 0)
))
}
if (bool_CI5_comp & is.null(age_label_list)) {
if (max_age < 18) {
dt_CI5[CSU_age_factor >= max_age , CSU_C:=sum(CSU_C)] ##add total_know
dt_CI5[ CSU_age_factor >= max_age , CSU_P:=sum(CSU_P)]
dt_CI5 <- dt_CI5[CSU_age_factor <= max_age]
}
dt_CI5$rate <- dt_CI5$CSU_C/dt_CI5$CSU_P *db_rate
if (logscale) {
dt_CI5[rate == 0 , rate:= NA]
}
}
ylim_inf <- min(c(tick$tick_list, tick$tick_minor_list))
ylim_sup <- max(c(tick$tick_list, tick$tick_minor_list))
##csu_plot
if (logscale) {
base_plot <- ggplot(dt_data[, rate := ifelse(rate==0,NA, rate )], aes(CSU_age_factor, rate))
} else {
base_plot <- ggplot(dt_data, aes(CSU_age_factor, rate))
}
if (bool_CI5_comp) {
pos_y_text = - tick_space
if (temp_top/tick_space > 7) {
pos_y_text = pos_y_text*1.5
}
if(is.null(plot_caption)) {
plot_caption <- paste0("- - - - - - : Mean for ",CI5_cancer_label," cancer in CI5 XII")
}
base_plot <- base_plot +
geom_line(data = dt_CI5,
linewidth = 1,
linetype=2,
colour = "grey50",
show.legend=FALSE)
}
csu_plot <- base_plot+
geom_line(aes(color=CSU_BY), linewidth = 1,na.rm=TRUE)+
guides(color = guide_legend(override.aes = list(size=0.75)))+
labs(title = plot_title,
subtitle = plot_subtitle,
caption = plot_caption)+
scale_x_continuous(name = xtitle,
breaks=seq(1, max_age, 1),
labels = age_label,
minor_breaks = NULL,
expand = c(0.015,0.015)
)
if (logscale){
if (log_point) {
csu_plot <- csu_plot +
geom_point(aes(fill=CSU_BY), size = 3,na.rm=TRUE,shape=21,stroke=0.5,colour="black", show.legend=FALSE)
}
csu_plot <- csu_plot +
scale_y_continuous(name = paste(ytitle, formatC(db_rate, format="d", big.mark=",")),
breaks=tick$tick_list,
minor_breaks = tick$tick_minor_list,
limits=c(ylim_inf,ylim_sup),
labels=core.csu_axes_label,
trans = "log10"
)
} else {
csu_plot <- csu_plot +
coord_cartesian( ylim=c(-temp_expand_y, temp_expand_y_up), expand = TRUE)+
scale_y_continuous(name = paste(ytitle, formatC(db_rate, format="d", big.mark=",")),
breaks=tick$tick_list,
labels=core.csu_axes_label,
expand = c(0,0)
)
}
csu_plot <- csu_plot +
theme(
plot.background= element_blank(),
panel.background = element_blank(),
panel.grid.major= element_line(colour = "grey70"),
panel.grid.minor= element_line(colour = "grey70"),
plot.title = element_text(size=16, margin=margin(0,0,15,0),hjust = 0.5),
plot.subtitle = element_text(size=15, margin=margin(0,0,15,0),hjust = 0.5),
plot.caption = element_text(size=10, margin=margin(15,0,0,0)),
axis.title = element_text(size=14),
axis.title.y = element_text(margin=margin(0,15,0,0)),
axis.title.x = element_text(margin=margin(15,0,0,0)),
plot.margin=margin(20,20,20,20),
axis.text = element_text(size=14, colour = "black"),
axis.text.x = element_text(size=14, angle = 60, hjust = 1),
axis.ticks= element_line(colour = "black", linewidth = linesize),
axis.ticks.length = unit(0.2, "cm"),
axis.line.x = element_line(colour = "black",
linewidth = linesize, linetype = "solid"),
axis.line.y = element_line(colour = "black",
linewidth = linesize, linetype = "solid")
)+
th_legend
if (!is.null(color_trend)) {
csu_plot <- csu_plot +
scale_colour_manual(name=legend$title,
labels = label_group_by,
values= color_trend,
drop = FALSE)
if (logscale) {
csu_plot <- csu_plot +
scale_fill_manual(labels = label_group_by,
values= color_trend,
drop = FALSE)
}
} else {
csu_plot <- csu_plot +
scale_colour_discrete(name=legend$title)
}
if (!bool_dum_by & legend$position=="right") {
csu_plot <- csu_plot +
geom_text(data = dt_data[CSU_age_factor == nb_age_group, ],
aes(label = CSU_BY),
hjust=-0.05)+
theme(plot.margin = unit(c(0.5, legend$right_space_margin, 0.5, 0.5), "lines"))
} else {
csu_plot <- csu_plot +
guides(color = guide_legend(nrow=legend$nrow))
}
dt_data$nb_age_group <- NULL
dt_data$CSU_age_factor <- NULL
if (logscale){
dt_data[, rate := ifelse(is.na(rate),0, rate )]
}
#get back age label
dt_data <- merge(dt_data, dt_temp_label, by=("CSU_A"), all.x=TRUE)
temp = grep("CSU_A", colnames(dt_data))
temp2 = length(colnames(dt_data))
dt_data <- dt_data[,c(1:temp, temp2, (temp+2):temp2-1) , with=FALSE ]
dt_data[, CSU_A := NULL]
setnames(dt_data, "temp_label", "CSU_A")
return(list(csu_plot = csu_plot, dt_data = dt_data, CI5_cancer_label = CI5_cancer_label,legend_position=legend$position,bool_dum_by = bool_dum_by))
}
core.csu_ageSpecific_top <- function(df_data,
var_age,
var_cases,
var_py,
var_top,
group_by=NULL,
missing_age=NULL,
db_rate = 100000,
logscale = FALSE,
nb_top = 5,
plot_title=NULL,
plot_subtitle=NULL,
label_group_by=NULL,
xtitle = "Age at diagnosis",
ytitle = "Age-specific incidence rate per",
var_color=NULL,
plot_caption=NULL,
var_age_label_list = NULL,
caption_bypass=FALSE) {
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
df_data$CSU_dum_by <- factor(df_data[[group_by]],levels=c("dummy_by"), labels=c(""))
bool_dum_by <- TRUE
} else {
df_data[[group_by]] <- as.factor(df_data[[group_by]])
}
df_data <- core.csu_dt_rank(df_data, var_value = var_cases, var_rank = var_top,group_by = group_by, number = nb_top)
df_data[["dummy_top"]] <-core.csu_legend_wrapper(df_data[[var_top]], 14)
plotlist <- list()
datalist <- list()
j <- 1
#dummmy variable to factorize variable
df_data$CSU_dum_by <- as.factor(df_data[[group_by]])
for (i in levels( df_data$CSU_dum_by)) {
if (!is.null(label_group_by)) {
label_group <- label_group_by[j]
}
else {
label_group <- i
}
if (caption_bypass) {
if (j == 1) {
plot_caption <- ""
} else {
plot_caption <- plot_title
plot_title <- ""
}
}
dt_plot <- df_data[get("CSU_dum_by") == i]
if (!is.null(var_color)) {
dt_label_order <- setkey(unique(dt_plot[, c("dummy_top",var_color, "CSU_RANK"), with=FALSE]), CSU_RANK)
dt_plot$dummy_top <- factor(dt_plot$dummy_top,levels = dt_label_order$dummy_top)
color_trend <- as.character(dt_label_order[[var_color]])
} else {
dt_label_order <- setkey(unique(dt_plot[, c(var_top, "CSU_RANK"), with=FALSE]), CSU_RANK)
dt_plot[[var_top]] <- factor(dt_plot[[var_top]],levels = dt_label_order[[var_top]])
color_trend <- NULL
}
if (!is.null(var_age_label_list)) {
age_label_list <- unique(dt_plot[[var_age_label_list]])
} else {
age_label_list <- NULL
}
if (is.null(plot_subtitle)) {
subtitle_temp <- label_group
} else {
subtitle_temp <- paste0(plot_subtitle,"\n",label_group)
}
temp <- core.csu_ageSpecific(
dt_plot,
var_age=var_age,
var_cases= var_cases,
var_py=var_py,
group_by = "dummy_top",
missing_age = missing_age,
db_rate= db_rate,
plot_title = plot_title,
plot_subtitle = subtitle_temp,
plot_caption = plot_caption,
color_trend = color_trend,
logscale = logscale,
log_point=FALSE,
age_label_list = age_label_list,
xtitle = xtitle,
ytitle = ytitle
)
dt_temp <- temp$dt_data
dt_temp[[group_by]] <- i
setnames(dt_temp, "CSU_BY", "dummy_top")
plotlist[[j]] <- temp$csu_plot
datalist[[j]] <- dt_temp
j <- j+1
}
df_data <- unique(df_data[,c("dummy_top",group_by, var_top, "CSU_RANK"), with=FALSE])
dt_final <- NULL
for (i in 1:(j-1)) {
dt_final <- rbind(dt_final, datalist[[i]])
}
dt_final <- merge(dt_final,df_data,by=c("dummy_top",group_by),all.x=TRUE, all.y=FALSE)
dt_final[, dummy_top :=NULL]
setnames(dt_final, "CSU_A", var_age)
setnames(dt_final, "CSU_C", var_cases)
setnames(dt_final, "CSU_P", var_py)
setkeyv(dt_final, c(group_by,"CSU_RANK",var_age))
if (bool_dum_by) {
dt_final[, CSU_dum_by:=NULL]
}
return(list(plotlist=plotlist, dt_data=dt_final))
}
core.csu_time_trend <- function (
df_data,
var_trend = "asr",
var_year = "year",
group_by = NULL,
logscale = TRUE,
smoothing = NULL,
legend = csu_trend_legend(),
color_trend = NULL,
ytitle = "Age standardized rate per 100000",
xtitle = "Year",
plot_title = "test",
linesize = 0.5,
plot_subtitle = NULL,
plot_caption = NULL) {
if (!is.null(smoothing)) {
if (smoothing == 0) {
smoothing <- NULL
}
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_year, "CSU_Y")
setnames(dt_data, var_trend, "CSU_T")
setnames(dt_data, group_by, "CSU_BY")
bool_dum_by <- FALSE
if (is.null(group_by)) {
dt_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
#change by to factor
dt_data$CSU_BY <- factor(dt_data$CSU_BY)
#smooth with loess fonction
if (!is.null(smoothing))
{
dt_data[,CSU_smooth:= loess( CSU_T ~ CSU_Y, span=smoothing)$fitted, by=CSU_BY]
} else {
dt_data[,CSU_smooth:= CSU_T]
}
dt_data[, max_year:=max(CSU_Y), by=CSU_BY]
# to calcul y axes breaks
if (logscale) {
min_tick_value <- min(dt_data[CSU_smooth > 0,]$CSU_smooth)
} else {
min_tick_value <- 0
}
tick <- core.csu_tick_generator(max = max(dt_data$CSU_smooth), min=min_tick_value, logscale = logscale )
tick_space <- tick$tick_list[length(tick$tick_list)] - tick$tick_list[length(tick$tick_list)-1]
#to calcul year axes break
year_tick <- core.csu_year_tick_generator(min(dt_data$CSU_Y),max(dt_data$CSU_Y))
temp_top <- ceiling(max(dt_data$CSU_smooth)/tick_space)*tick_space
temp_expand_y <- max(dt_data$CSU_smooth)/35
temp_expand_y_up <- max(dt_data$CSU_smooth)+temp_expand_y
if (temp_expand_y_up > temp_top-(tick_space/2)) {
temp_expand_y_up <- temp_top+temp_expand_y
}
th_legend <- list(theme(legend.position="none"))
if (!bool_dum_by & legend$position == "bottom") {
th_legend <- list(theme(
legend.key = element_rect(fill="transparent"),
legend.position = "bottom",
legend.text = element_text(size = 12),
legend.title = element_text(size = 12),
legend.key.size=unit(1,"cm"),
legend.margin = margin(0, 0, 0, 0)
))
}
xlim_inf <- min(c(year_tick$tick_list, year_tick$tick_minor_list))
xlim_sup <- max(c(year_tick$tick_list, year_tick$tick_minor_list))
ylim_inf <- min(c(tick$tick_list, tick$tick_minor_list))
ylim_sup <- max(c(tick$tick_list, tick$tick_minor_list))
#csu_plot
if (logscale) {
base_plot <- ggplot(dt_data[, CSU_smooth := ifelse(CSU_smooth==0,NA, CSU_smooth )], aes(CSU_Y, CSU_smooth))
} else {
base_plot <- ggplot(dt_data, aes(CSU_Y, CSU_smooth))
}
csu_plot <- base_plot+
geom_line(aes(color=CSU_BY), linewidth = 0.75,na.rm=TRUE)+
guides(color = guide_legend(override.aes = list(size=0.75)))+
labs(title = plot_title,
subtitle = plot_subtitle,
caption = plot_caption)+
scale_x_continuous(name = xtitle,
breaks=year_tick$tick_list,
limits=c(xlim_inf,xlim_sup),
minor_breaks = year_tick$tick_minor_list,
expand = c(0.015,0.015)
)
if (logscale){
csu_plot <- csu_plot +
scale_y_continuous(name = ytitle,
breaks=tick$tick_list,
minor_breaks = tick$tick_minor_list,
limits=c(ylim_inf,ylim_sup),
labels=core.csu_axes_label,
trans = "log10"
)
} else {
csu_plot <- csu_plot +
coord_cartesian( ylim=c(-temp_expand_y, temp_expand_y_up), expand = TRUE)+
scale_y_continuous(name = ytitle,
breaks=tick$tick_list,
labels=core.csu_axes_label,
expand = c(0,0)
)
}
if (is.null(color_trend)) {
csu_plot <- csu_plot +scale_colour_discrete(name=legend$title)
}
else {
csu_plot <- csu_plot +scale_colour_manual(name=NULL, values=color_trend)
}
csu_plot <- csu_plot +
theme(
plot.background= element_blank(),
panel.background = element_blank(),
panel.grid.major= element_line(colour = "grey70"),
panel.grid.minor= element_line(colour = "grey70"),
plot.title = element_text(size=16, margin=margin(0,0,15,0),hjust = 0.5),
plot.subtitle = element_text(size=15, margin=margin(0,0,15,0),hjust = 0.5),
plot.caption = element_text(size=10, margin=margin(15,0,0,0)),
axis.title = element_text(size=12),
axis.title.y = element_text(margin=margin(0,15,0,0)),
axis.title.x = element_text(margin=margin(15,0,0,0)),
plot.margin=margin(20,20,20,20),
axis.text = element_text(size=12, colour = "black"),
axis.text.x = element_text(size=12, hjust = 0.5),
axis.ticks= element_line(colour = "black", linewidth = linesize),
axis.ticks.length = unit(0.2, "cm"),
axis.line.x = element_line(colour = "black",
linewidth = linesize, linetype = "solid"),
axis.line.y = element_line(colour = "black",
linewidth = linesize, linetype = "solid")
)+
th_legend
if (!bool_dum_by & legend$position=="right") {
csu_plot <- csu_plot +
geom_text(data = dt_data[CSU_Y == max_year, ],
aes(label = CSU_BY),
hjust=0,
nudge_x=0.5)+
theme(plot.margin = unit(c(0.5, legend$right_space_margin, 0.5, 0.5), "lines"))
} else {
csu_plot <- csu_plot +
guides(color = guide_legend(nrow=legend$nrow))
}
return(list(dt_data=dt_data, csu_plot = csu_plot, legend_position=legend$position,bool_dum_by = bool_dum_by))
}
core.csu_year_extract <- function(year_list) {
temp <- gsub("[^\\d]", "", year_list, perl=TRUE)
if (unique(nchar(temp)) == 4 ) {
return(as.numeric(year_list))
}
else if (unique(nchar(temp)) == 6 ) {
test <- all(grepl("(18|19|20)\\d{2}\\d{2}",temp))
if (test) {
return(as.numeric(gsub("(^\\d{4}).+", "\\1", temp, perl=TRUE)))
}
else {
return(as.numeric(gsub(".+(\\d{4}$)", "\\1", temp, perl=TRUE)))
}
}
else if (unique(nchar(temp)) == 8 ) {
test <- all(grepl("(18|19|20)\\d{2}\\d{4}",temp))
if (test) {
return(as.numeric(gsub("(^\\d{4}).+", "\\1", temp, perl=TRUE)))
}
else {
return(as.numeric(gsub(".+(\\d{4}$)", "\\1", temp, perl=TRUE)))
}
}
}
core.csu_icd_group <- function(icd_list) {
bool_follow <- FALSE
icd_first <- icd_list[1]
icd_long <- icd_first
code_active = as.numeric(sub(".+?(\\d+)", "\\1", icd_first))
for (code in icd_list[-1]) {
code_new = as.numeric(sub(".+?(\\d+)", "\\1", code))
bool_follow <- code_new == code_active + 1
if (bool_follow) {
icd_long <- paste0(icd_first, '-', sprintf("%02d", code_new))
}
else {
icd_long <- paste0(icd_long, ',', as.character(code))
icd_first <- icd_long
}
code_active <- code_new
}
return(icd_long)
}
core.csu_icd_ungroup <- function(icd_group) {
icd_group <- gsub("\\s", "", icd_group)
icd_list <- NULL
ICD_reg <-"([A-Za-z]\\d+)(\\W?)(.+)?"
while (nchar(icd_group)>=3) {
icd_start <- sub(ICD_reg, "\\1", icd_group)
icd_mark <- sub(ICD_reg, "\\2", icd_group)
icd_group <- sub(ICD_reg, "\\3", icd_group)
if (icd_mark == "-") {
letter_start <- sub("([A-Za-z])(\\d+)", "\\1", icd_start)
code_start <- sub("([A-Za-z])(\\d+)", "\\2", icd_start)
code_nchar <- nchar(code_start)
code_start <- as.numeric(code_start)
code_end <- as.numeric(sub("[A-Za-z]?(\\d+)(.+)?", "\\1", icd_group))
for (code in code_start:code_end) {
icd_list <- c(icd_list, paste0(letter_start, sprintf(paste0("%0",code_nchar,"d"), code)))
}
icd_group <- sub("([A-Za-z]?\\d+)(\\W?)(.+)?", "\\3", icd_group)
}
else {
icd_list <- c(icd_list, sub("([A-Za-z]\\d+)", "\\1", icd_start))
}
}
return(icd_list)
}
core.csu_group_cases <- function(df_data, var_age ,group_by=NULL,var_cases = NULL,df_ICD = NULL,var_ICD=NULL,var_year = NULL, all_cancer=FALSE) {
label_by <- NULL
dt_data <- data.table(df_data)
bool_dum_by <- FALSE
if (is.null(group_by)) {
dt_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
if (is.null(var_cases)) {
var_cases <- "CSU_C"
dt_data[, CSU_C:= 1]
} else {
setnames(dt_data, var_cases, "CSU_C")
}
dt_data <- dt_data[, unique(c("CSU_C", var_age,group_by,var_ICD,var_year)), with = FALSE]
if (!is.null(var_year)) {
dt_data$year <- core.csu_year_extract(dt_data[[var_year]])
if (var_year != "year") {
dt_data[, (var_year) := NULL]
}
group_by <- c(group_by, "year")
}
if (!is.null(df_ICD)) {
# merge with ICD
dt_ICD <- data.table(df_ICD)
setkeyv(dt_ICD,c("LABEL", "ICD"))
dt_ICD <- unique(dt_ICD)
#ungroup ICD code
dt_table <- data.table()
for (row in 1:nrow(dt_ICD)) {
icd_group <- as.character(dt_ICD[row]$ICD)
temp <- core.csu_icd_ungroup(paste(icd_group, collapse=","))
temp <- data.table(ICD_ungroup = temp, ICD =icd_group )
dt_table <- rbind(dt_table, temp)
}
dt_ICD <- merge(dt_ICD, dt_table, by="ICD")
dt_ICD[, ICD:=NULL]
dt_ICD <- unique(dt_ICD)
dt_ICD_unique <- setDT(dt_ICD)[, .N, keyby=ICD_ungroup][N>1,]
if (nrow(dt_ICD_unique) > 0) {
dt_ICD_unique <- merge(dt_ICD_unique,dt_ICD,by="ICD_ungroup", all.x=TRUE)
dt_ICD_unique <- merge(dt_ICD_unique, df_ICD, by="LABEL")
dt_ICD_unique <- dt_ICD_unique[,c("LABEL", "ICD") , with=FALSE]
cat("-----\n")
print(as.data.frame(dt_ICD_unique))
stop(paste0("There is different label with the same ICD code in the ICD file defined"))
}
setnames(dt_ICD, "ICD_ungroup", "ICD")
dt_ICD[, ICD_group:= sapply(LABEL, function(x) {core.csu_icd_group(as.vector(dt_ICD[LABEL == x, ]$ICD))})]
list_ICD <- dt_ICD$ICD
dt_data$temp <- as.character(dt_data[[var_ICD]])
dt_data[, ICD := list_ICD[match(dt_data$temp, list_ICD)]]
dt_data[!is.na(ICD), temp := NA]
dt_data[, temp:= substr(temp, 1, 3)]
dt_data[!is.na(temp), ICD:=list_ICD[match(dt_data[!is.na(temp)]$temp, list_ICD)]]
dt_data <- dt_data[!is.na(ICD), ]
dt_data <- merge(dt_data, dt_ICD, by=c("ICD"))
dt_data[,c("temp", var_ICD,"ICD") := list(NULL, NULL,NULL)]
dt_data[, LABEL := factor(LABEL)]
dt_data[, ICD_group := factor(ICD_group)]
label_by <- c(label_by,"LABEL")
group_by <- c(group_by, "ICD_group")
dt_ICD <- NULL
}
# create age group
dt_data[, age_group:= ((get(var_age) - (get(var_age) %% 5))/5)+1]
dt_data[, age_group := ifelse(age_group > 18 & age_group <31,18,age_group)]
dt_data[, age_group := ifelse(age_group > 18 ,19,age_group)]
dt_data[, c(var_age) := NULL]
dt_data <- dt_data[,list(CSU_C = sum(CSU_C)),by=eval(colnames(dt_data)[!colnames(dt_data) %in% c("CSU_C")])]
#complete missing age group
if (max(dt_data$age_group) == 19) {
temp <- c(1:19)
}
else {
temp <- c(1:18)
}
dt_CJ = dt_data[, do.call(CJ, c(.SD,list(age_group=temp), unique=TRUE)), .SDcols=group_by]
temp <- copy(colnames(dt_CJ))
# add age_group label
dt_CJ[, temp1 := sprintf("%02d",(age_group-1)*5)]
dt_CJ[, temp2 := (age_group*5)-1]
dt_CJ[, age_group_label := ifelse(temp2 == 89, paste0(temp1,"+"), paste0(temp1,"-", sprintf("%02d", temp2)))]
dt_CJ[age_group == 19 , age_group_label := "Unknown"]
dt_CJ[,c("temp1","temp2") := list(NULL, NULL)]
group_by <- c(group_by, "age_group")
##add ICD group label (but #dad is almost pink)
if (!is.null(df_ICD)) {
dt_temp <- unique(dt_data[, c("ICD_group","LABEL"), with=FALSE])
dt_CJ <- merge(dt_CJ, dt_temp, by="ICD_group", all.x=TRUE)
temp <- c(temp, "LABEL")
}
dt_data <- merge(dt_CJ, dt_data,by=temp, all.x=TRUE)[, CSU_C := ifelse(is.na(CSU_C),0, CSU_C )]
keep_by <- group_by[!group_by %in% c("year", "age_group", "ICD_group")]
setkeyv(dt_data, keep_by)
dt_data[ , group_id := .GRP, by = key(dt_data)]
dt_data[,temp := sum(CSU_C),by=group_id]
dt_data <- dt_data[temp>0,]
dt_data[, c("temp","group_id") := NULL]
setkeyv(dt_data, group_by)
if (var_cases=="CSU_C") {
setnames(dt_data, "CSU_C","cases")
var_cases <- NULL
}
else {
setnames(dt_data, "CSU_C",var_cases)
}
if (bool_dum_by) {
dt_data$CSU_dum_by <- NULL
}
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
return (df_data)
}
Try the Rcan package in your browser
Any scripts or data that you put into this service are public.
Rcan documentation built on Oct. 7, 2024, 1:18 a.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 core.csu_group_cases core.csu_icd_ungroup core.csu_icd_group core.csu_year_extract core.csu_ageSpecific_top core.csu_ageSpecific core.csu_eapc core.csu_cumrisk core.csu_asr core.csu_format_export core.csu_year_tick_generator core.csu_axes_label core.csu_tick_generator core.csu_legend_wrapper core.csu_dt_rank core.error_time_variable core.error_age_parse core.error_missingage core.error_variable
core.error_variable <- function(df_data, varname, funcname,type="numeric") {
if (!is.null(varname)) {
if (!(varname%in% colnames(df_data))) {
stop(paste0(varname, " is not a column of ", deparse(substitute(df_data)), " ,see documentation: Help(", deparse(substitute(funcname)), ")"))
}
if (type == "numeric") {
if (!is.numeric(df_data[[varname]])) {
stop(paste0(varname, " type must be ",type," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
} else if (type == "character"){
if (!is.character(df_data[[varname]])) {
stop(paste0(varname, " type must be ",type," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
} else if (type == "factor"){
if (!is.factor(df_data[[varname]])) {
stop(paste0(varname, " type must be ",type," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
}
}
}
core.error_missingage <- function(df_data, var_age,missing_age,funcname) {
if (!missing_age %in% df_data[[var_age]]) {
stop(paste0("\n",missing_age, " value is not present in the variable ",var_age," see documentation: Help(", deparse(substitute(funcname)), ")"))
}
}
core.error_age_parse <- function(df_data, var_age, missing_age,funcname) {
dt_data <- as.data.table(df_data)
dt_data <- dt_data[, c(var_age), with=FALSE]
setnames(dt_data, var_age, 'CSU_A')
if (!is.null(missing_age)) {
dt_data[CSU_A == missing_age, CSU_A := NA]
}
if (!all(grepl(".*?(\\d{1,3}).*$",dt_data[!is.na(CSU_A)]$CSU_A, perl=TRUE)))
{
temp <- unique(as.character(dt_data[!grepl(".*?(\\d{1,3}).*$",dt_data[!is.na(CSU_A)]$CSU_A, perl=TRUE),]$CSU_A))
stop(paste0("\nthe value ",temp,", in the variable ",var_age, "\ncannot be change to numeric and is not declared as missing age,\nSee documentation: Help(", deparse(substitute(funcname)), ")\n"))
}
}
core.error_time_variable <- function(df_data, var_year, group_by, funcname) {
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
}
dt_data <- data.table(df_data, key = c(group_by))
dt_data$temp <- 1
nrow_base <- nrow(dt_data)
dt_test <- dt_data[ ,temp:=sum(temp), by=c(var_year, group_by)]
nrow_test <- nrow(dt_data[ ,sum(temp), by=c(var_year, group_by)])
if (nrow_test != nrow_base) {
setkeyv(dt_test, c(group_by,var_year))
print(head(dt_test[temp>1, ]))
dt_data <- NULL
stop(paste0("There is more than 1 data per year (see above).\nUse the 'group_by' option"," ,see documentation: Help(", deparse(substitute(funcname)), ")\n","or call the function on a subset to define the sub-population of interest.\n"))
}
}
core.csu_dt_rank <- function(df_data,
var_value = "CASES",
var_rank = "cancer_label",
group_by = NULL,
number = NULL,
ties.method="min") {
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
df_data <- as.data.table(df_data)
dt_rank <- df_data[, list(rank_value=sum(get(var_value))), by=c(var_rank, group_by)]
dt_rank[, CSU_RANK:= frank(-rank_value, ties.method=ties.method), by=group_by]
if (!is.null(number)){
dt_rank <- dt_rank[CSU_RANK <= number,c(group_by, var_rank, "CSU_RANK"), with=FALSE]
}
df_data <- merge(dt_rank, df_data,by=c(group_by, var_rank), all.x=TRUE)
if (bool_dum_by) {
df_data[,CSU_dum_by:=NULL]
}
return(df_data)
}
core.csu_legend_wrapper <- function(label, width) {
label <- sapply(strwrap(label, width = width, simplify = FALSE), paste, collapse="\n")
return(label)
}
core.csu_tick_generator <- function(max,min = 0,logscale=FALSE) {
if (!logscale) {
if (min > 0) {
min = 0
}
if (max < 0) {
max = 0
}
temp_log_max = 10^floor(log10(max-min))
temp_unit_floor_max = floor((max-min)/(temp_log_max))
if (temp_unit_floor_max < 2) {
tick_space = 0.2*temp_log_max
} else {
if (temp_unit_floor_max < 5) {
tick_space = 0.5*temp_log_max
} else {
tick_space = temp_log_max
}
}
temp_top <- ceiling(max/tick_space)*tick_space
temp_floor <- floor(min/tick_space)*tick_space
tick_list <- seq(temp_floor, temp_top, tick_space)
tick_minor_list <- NULL
} else {
temp_log_max = 10^floor(log10(max))
temp_unit_floor_max = floor(max/(temp_log_max))
temp_log_min <- 10^floor(log10(min))
temp_unit_floor_min <- floor(min/(temp_log_min))
if (temp_log_min == temp_log_max) {
tick_list <- c(temp_unit_floor_min:(temp_unit_floor_max+1)*temp_log_min)
if (temp_unit_floor_max == temp_unit_floor_min) {
tick_minor_list <- c((temp_unit_floor_min*temp_log_min)+0:9*(temp_log_min/10))
} else {
tick_minor_list <- c((temp_unit_floor_min*temp_log_min)+0:19*(temp_log_min/10))
}
} else if (temp_log_max/temp_log_min < 1000) {
if (temp_unit_floor_min < 6) {
tick_list <- temp_unit_floor_min:5*temp_log_min
tick_list <- c(tick_list, temp_log_min*7) ## min . . 5 7
} else {
tick_list <- temp_unit_floor_min*temp_log_min ## min
}
tick_minor_list <- temp_unit_floor_min:19*temp_log_min ## min . . 19
while (temp_log_min != (temp_log_max/10)) {
temp_log_min = temp_log_min*10
tick_list <- c(tick_list, c(1,2,3,5,7)*temp_log_min)
tick_minor_list <- c(tick_minor_list, 2:19*temp_log_min)
}
tick_minor_list <- c(tick_minor_list, 2:(temp_unit_floor_max+1)*temp_log_max)
if (temp_unit_floor_max <5) {
tick_list <- c(tick_list, 1:(temp_unit_floor_max+1)*temp_log_max)
} else if (temp_unit_floor_max <7) {
tick_list <- c(tick_list, c(1,2,3,5,temp_unit_floor_max+1)*temp_log_max)
} else {
tick_list <- c(tick_list, c(1,2,3,5,7,temp_unit_floor_max+1)*temp_log_max)
}
} else {
if (temp_unit_floor_min == 1) {
tick_list <- c(1,2,3,5)*temp_log_min
} else if (temp_unit_floor_min == 2) {
tick_list <- c(2,3,5)*temp_log_min
} else if (temp_unit_floor_min < 6) {
tick_list <- c(5,7)*temp_log_min
} else {
tick_list <- 7*temp_log_min
}
tick_minor_list <- temp_unit_floor_min:9*temp_log_min ## min . . 19
while (temp_log_min != (temp_log_max/10)) {
temp_log_min = temp_log_min*10
tick_list <- c(tick_list, c(1,2,5)*temp_log_min)
tick_minor_list <- c(tick_minor_list, 2:9*temp_log_min)
}
tick_minor_list <- c(tick_minor_list, 2:(temp_unit_floor_max+1)*temp_log_max)
if (temp_unit_floor_max <5) {
tick_list <- c(tick_list, unique(c(1,2,temp_unit_floor_max+1)*temp_log_max))
} else if (temp_unit_floor_max <6) {
tick_list <-c(tick_list, c(1,2,5)*temp_log_max)
} else if (temp_unit_floor_max < 7) {
tick_list <- c(tick_list, c(1,2,5,7)*temp_log_max)
} else {
tick_list <- c(tick_list, c(1,2,5,7,temp_unit_floor_max+1)*temp_log_max)
}
}
}
return(list(tick_list=tick_list, tick_minor_list=tick_minor_list))
}
core.csu_axes_label <- function(l) {
l <- format(l, big.mark = ",", scientific = FALSE, drop0trailing = TRUE)
}
core.csu_year_tick_generator <- function(min, max) {
mod <- 5
if (max - min < 10 ) {
mod <- 1
} else if (max - min < 20){
mod <- 2
}
temp1 <- min - (min %% mod)
temp2 <- max - (max %% mod) +ifelse(mod>=5,mod,0)
if (temp2 - temp1 <= mod*6) {
year_space <- mod
year_list <- seq(temp1,temp2,year_space)
year_minor_list <- year_list
} else {
year_space <- mod*2
if (temp1 %% mod*2 > 0) {
year_list <- seq(temp1+mod,temp2,year_space)
year_minor_list <- seq(temp1,temp2,year_space/2)
} else {
year_list <- seq(temp1,temp2,year_space)
year_minor_list <- seq(temp1,temp2,year_space/2)
}
}
return(list(tick_list=year_list, tick_minor_list=year_minor_list))
}
core.csu_format_export <- function(type, plot_title, landscape=FALSE) {
#http://www.altelia.fr/actualites/calculateur-resolution-definition-format.htm
# 6 inch = 15.24 cm
#10,795
#format
#if (!is.null(format_export)) {
# filename <- gsub("[[:punct:][:space:]\n]", "_", plot_title)
# core.csu_format_export(format_export, plot_title = filename, landscape = FALSE)
#}
# help file
# \item{format_export}{
# export the graph in different format:\cr
# \tabular{ll}{
# \code{NULL} \tab Plot in R studio windows.\cr
# \code{"pdf"} \tab Export in PDF format.\cr
# \code{"tiff"} \tab Export in TIFF 300dpi format.\cr
# \code{"png"} \tab Export in PNG 200dpi format.\cr
# \code{"svg"} \tab Export in SVG format. Can be edit with \url{https://inkscape.org/fr/}.\cr
# }
# The filename is the \code{plot_title} option.
# }
# \item{graph_dev}{
# If the plot is embedded in a graphics Device function (such as pdf()), the graph_dev option should be set to TRUE for the first graph to avoid a blank page.
# }
png_width <- ifelse(landscape, 2339 , 1654 )
png_height <- ifelse(landscape, 1654 , 2339 )
tiff_width <- ifelse(landscape, 3508 , 2480 )
tiff_height <- ifelse(landscape, 2480 , 3508 )
pdf_width <- ifelse(landscape, 11.692 , 8.267 )
pdf_height <- ifelse(landscape, 8.267 , 11.692 )
if (type == "pdf") {
pdf(paste(plot_title,".",type, sep=""),width = pdf_width, height = pdf_height)
} else if (type == "svg") {
svg(paste(plot_title,".",type, sep=""),width = pdf_width, height = pdf_height)
} else if (type == "png") {
png(paste(plot_title,".",type, sep=""),width = png_width, height = png_height, units = "px",res = 200)
} else if (type == "tiff") {
tiff(paste(plot_title,".",type, sep=""),width = tiff_width, height = tiff_height,units = "px",res = 300,compression ="lzw")
}
}
core.csu_asr <- function(df_data, var_age, var_cases, var_py, group_by=NULL,
var_age_group=NULL, missing_age = NULL,db_rate = 100000,
first_age = 1, last_age = 18, pop_base = "SEGI", crude_rate = FALSE,
var_st_err=NULL,correction_info=FALSE, var_asr="asr",age_dropped = FALSE,
pop_base_count = NULL, age_label_list = NULL, Rcan_print=FALSE)
{
bool_dum_by <- FALSE
bool_dum_age <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
if (is.null(var_age_group)) {
df_data$CSU_dum_age <- "dummy_age_gr"
var_age_group <- "CSU_dum_age"
group_by <- c(group_by, "CSU_dum_age")
bool_dum_age <- TRUE
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_age, "CSU_A")
setnames(dt_data, var_cases, "CSU_C")
setnames(dt_data, var_py, "CSU_P")
if (crude_rate == FALSE)
{
if (is.null(pop_base_count)) {
dt_data <- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P)), by=c(group_by, "CSU_A")]
}
# create index to keep order
index_order <- c(1:nrow(dt_data))
dt_data$index_order <- index_order
# missing age
dt_data[dt_data$CSU_A %in% missing_age,CSU_A:=NA ]
dt_data[is.na(dt_data$CSU_A),CSU_P:=0 ]
#parse age
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
#create age dummy: 1 2 3 4 --- 19
dt_data$age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
# correction factor
dt_data$correction <- 1
if (!is.null(missing_age)) {
dt_data[, total:=sum(CSU_C), by=group_by] #add total
dt_data[!is.na(dt_data$age_factor) , total_known:=sum(CSU_C), by=group_by] #add total_know
dt_data$correction <- dt_data$total / dt_data$total_know
dt_data[is.na(dt_data$correction),correction:=1 ]
dt_data$total <- NULL
dt_data$total_known <- NULL
}
if (is.null(pop_base_count)) {
# create world population DF for different nb of age group
SEGI_pop <- c(12000,10000,9000,9000,8000,8000,6000,6000,6000,6000,5000,4000,4000,3000,2000,1000,500,500)
EURO_pop <- c(8000,7000,7000,7000,7000,7000,7000,7000,7000,7000,7000,6000,5000,4000,3000,2000,1000,1000)
EURO2_pop <- c(5000,5500,5500,5500,6000,6000,6500,7000,7000,7000,7000,6500,6000,5500,5000,4000,2500,2500)
WHO_pop <- c(8860,8690,8600,8470,8220,7930,7610,7150,6590,6040,5370,4550,3720,2960,2210,1520,910,630)
if (pop_base == "EURO") {
pop <- EURO_pop
} else if (pop_base == "EURO2") {
pop <- EURO2_pop
} else if (pop_base == "WHO") {
pop <- WHO_pop
} else {
pop <- SEGI_pop
}
# calculated total pop for age selected
total_pop <- sum(pop[first_age:last_age])
Standard_pop <- data.table(pop = pop, age_factor= c(1:18))
pop[17] <- pop[17]+ pop[18]
pop[18] <- 0
Standard_pop$pop17 <- pop
pop[16] <- pop[16]+ pop[17]
pop[17] <- 0
Standard_pop$pop16 <- pop
pop[15] <- pop[15]+ pop[16]
pop[16] <- 0
Standard_pop$pop15 <- pop
#age dropped option
if (age_dropped) {
dt_data$age_factor <- dt_data$age_factor + first_age -1
}
# keep age selected
dt_data=dt_data[dt_data$age_factor %in% c(first_age:last_age) | is.na(dt_data$age_factor), ]
# calculated maximum age group with population data
if (last_age == 18) {
dt_data <- merge(dt_data, dt_data[dt_data$CSU_P != 0,list(nb_age_group = max(age_factor)), by=var_age_group], by=var_age_group)
} else {
dt_data$nb_age_group <- 18
}
# show population with less than 18 age group
if (last_age == 18) {
temp <- subset(dt_data,nb_age_group <18, select= c(var_age_group, "nb_age_group"))
if (nrow(temp) >0) {
setkey(temp,NULL)
if (Rcan_print) {
cat("\n")
cat("Population with less than 18 age group:\n\n" )
print
print(unique(temp), row.names = FALSE)
cat("\n")
}
}
temp <- NULL
}
#regroup case for population with nb of age group < 18
for (i in 15:17) {
if (i %in% dt_data$nb_age_group) {
dt_data[nb_age_group == i & age_factor >= i , CSU_C:=sum(CSU_C), by=group_by] #add total_know
dt_data[nb_age_group == i & age_factor > i & !is.na(age_factor), CSU_C := 0]
}
}
#add world pop to database
dt_data <- merge(dt_data,Standard_pop, by =c("age_factor"), all.x=TRUE )
Standard_pop <- NULL
dt_data[nb_age_group==17, pop:=pop17]
dt_data[nb_age_group==16, pop:=pop16]
dt_data[nb_age_group==15, pop:=pop15]
#return(dt_data)
} else {
#keep age group selected
dt_data <- dt_data[age_factor %in% (first_age:last_age), ]
#calcul total pop for canreg
total_pop <-sum(unique(dt_data[, c("age_factor", pop_base_count), with=FALSE])[[pop_base_count]])
#get age group list variable
if (is.null(age_label_list)) {
age_label_list <- var_age
}
age_group_list <- as.character(unique(dt_data[[age_label_list]]))
age_group_list <- paste(age_group_list, collapse=" ")
#rename variable population reference
setnames(dt_data, pop_base_count, "pop")
}
#calcul ASR
dt_data[dt_data$CSU_P != 0,rate:= dt_data$CSU_C[dt_data$CSU_P != 0]/ dt_data$CSU_P[dt_data$CSU_P != 0] * db_rate]
dt_data$asr <- dt_data$rate * dt_data$pop
dt_data[is.na(dt_data$asr),asr:=0 ]
dt_data$st_err <- ( dt_data$rate * (dt_data$pop^2) * (db_rate - dt_data$rate))/dt_data$CSU_P
dt_data[is.na(dt_data$st_err),st_err:=0 ]
# to check order
dt_data<- dt_data[order(dt_data$index_order ),]
dt_data<- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P),asr=sum(asr),st_err = sum(st_err),correction = max(correction)), by=group_by]
dt_data$asr <- dt_data$asr / total_pop
dt_data$asr <- dt_data$asr * dt_data$correction
dt_data$st_err <- (dt_data$st_err / (total_pop^2))^(1/2)
dt_data$st_err <- dt_data$st_err * dt_data$correction
dt_data$asr <- round(dt_data$asr, digits = 2)
dt_data$st_err <- round(dt_data$st_err, digits = 2)
dt_data$correction <- round((dt_data$correction-1)*100, digits = 1)
if (is.null(var_st_err)) {
dt_data$st_err <- NULL
} else {
setnames(dt_data, "st_err", var_st_err)
}
if (var_asr!="asr") {
setnames(dt_data, "asr", var_asr)
}
if (!correction_info) {
dt_data$correction <- NULL
}
}
else
{
#parse age
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
#create age dummy: 1 2 3 4 --- 19
dt_data$age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
#age dropped option
if (age_dropped) {
dt_data$age_factor <- dt_data$age_factor + first_age -1
}
# keep age selected
dt_data=dt_data[dt_data$age_factor %in% c(first_age:last_age) | is.na(dt_data$age_factor), ]
dt_data <- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P)), by=c(group_by)]
dt_data <- dt_data[, CSU_RESULT := CSU_C/CSU_P*db_rate]
if (var_asr=="asr") {
setnames(dt_data, "CSU_RESULT", 'crude_rate')
}
else
{
setnames(dt_data, "CSU_RESULT", var_asr)
}
}
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
if (bool_dum_age) {
df_data$CSU_dum_age <- NULL
}
if (bool_dum_by) {
df_data$CSU_dum_by <- NULL
}
setnames(df_data, "CSU_C", var_cases)
setnames(df_data, "CSU_P", var_py)
if (is.null(pop_base_count)) {
if (Rcan_print) {
temp <- last_age*5-1
if (last_age == 18) temp <- "99+"
if (crude_rate == FALSE)
{
cat("ASR have been computed for the age group ", (first_age-1)*5,"-", temp , "\n", sep="" )
}
else {
cat("Crude rate have been computed for the age group ", (first_age-1)*5,"-", temp , "\n", sep="" )
}
}
temp<- NULL
} else {
#cat("ASR have been computed for the age groups:\n",age_group_list , "\n", sep="" )
age_group_list<- NULL
}
return(df_data)
}
core.csu_cumrisk <- function(df_data, var_age, var_cases, var_py, group_by=NULL,
missing_age = NULL, last_age = 18,var_st_err=NULL,correction_info=FALSE,
var_cumrisk="cumrisk", age_label_list = NULL, Rcan_print=FALSE)
{
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_age, "CSU_A")
setnames(dt_data, var_cases, "CSU_C")
setnames(dt_data, var_py, "CSU_P")
dt_data <- dt_data[,list( CSU_C=sum(CSU_C), CSU_P=sum(CSU_P)), by=c(group_by, "CSU_A")]
# create index to keep order
index_order <- c(1:nrow(dt_data))
dt_data$index_order <- index_order
# missing age
dt_data[dt_data$CSU_A %in% missing_age,CSU_A:=NA ]
dt_data[is.na(dt_data$CSU_A),CSU_P:=0 ]
#parse age
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
#create age dummy: 1 2 3 4 --- 19
dt_data$age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
# correction factor
dt_data$correction <- 1
if (!is.null(missing_age)) {
dt_data[, total:=sum(CSU_C), by=group_by] #add total
dt_data[!is.na(dt_data$age_factor) , total_known:=sum(CSU_C), by=group_by] #add total_know
dt_data$correction <- dt_data$total / dt_data$total_know
dt_data[is.na(dt_data$correction),correction:=1 ]
dt_data$total <- NULL
dt_data$total_known <- NULL
dt_data<- dt_data[!is.na(age_factor),]
}
if (!is.null(age_label_list)) {
# calcul year interval from age group label
dt_temp <- unique(dt_data[, c(age_label_list), with=FALSE])
dt_temp[, min:=as.numeric(regmatches(get(age_label_list), regexpr("[0-9]+",get(age_label_list))))]
dt_temp[, max:=shift(min, type ="lead")]
dt_temp[, age_span := max-min]
dt_temp <- dt_temp[, c("age_span",age_label_list), with=FALSE]
dt_data <- merge(dt_data, dt_temp,by= age_label_list, all.x=TRUE)
} else {
dt_data[, age_span:=5]
}
#keep age group selected
age_max <- max(dt_data$age_factor)
if (age_max-1 < last_age) {
last_age <- age_max-1
}
dt_data=dt_data[dt_data$age_factor <= last_age]
#calculate cum risk
dt_data[,cumrisk:=age_span*(CSU_C/CSU_P)]
dt_data[CSU_P==0,cumrisk:=0]
dt_data[, st_err := (CSU_C/CSU_P)/CSU_P]
dt_data[is.na(dt_data$st_err),st_err:=0 ]
# to check order
dt_data<- dt_data[order(dt_data$index_order ),]
dt_data <- dt_data[,list( cumrisk=sum(cumrisk), CSU_P=sum(CSU_P),CSU_C=sum(CSU_C),st_err = sum(st_err),correction = max(correction)), by=group_by]
dt_data[,cumrisk:=(1-exp(-cumrisk))*100*correction]
dt_data[, st_err:=(st_err^(1/2))*100*5]
dt_data[, st_err:=st_err*correction]
dt_data[,cumrisk:=round(cumrisk, digits = 2)]
dt_data[,st_err:=round(st_err, digits = 2)]
dt_data[, correction:=round((correction-1)*100, digits = 1)]
if (is.null(var_st_err)) {
dt_data$st_err <- NULL
} else {
setnames(dt_data, "st_err", var_st_err)
}
if (var_cumrisk!="cumrisk") {
setnames(dt_data, "cumrisk", var_cum_risk)
}
if (!correction_info) {
dt_data$correction <- NULL
}
setnames(dt_data, "CSU_C", var_cases)
setnames(dt_data, "CSU_P", var_py)
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
if (bool_dum_by) {
df_data$CSU_dum_by <- NULL
}
if (Rcan_print) {
temp <- last_age*5-1
cat("Cumulative risk have been computed for the age group ","0-", temp , "\n", sep="" )
temp<- NULL
}
return(df_data)
}
core.csu_eapc <- function(df_data,
var_rate="asr",
var_year="year",
group_by= NULL,
var_eapc="eapc",
CI_level = 0.95)
{
#create fake group to have group_by optional
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
dt_data <- data.table(df_data, key = c(group_by))
setnames(dt_data, var_rate, "CSU_R")
setnames(dt_data, var_year, "CSU_Y")
# remove 0 data
dt_data <- dt_data[!CSU_R == 0,]
dt_data[, id_group:=.GRP, by=group_by]
temp_max <- max(dt_data$id_group)
for (i in 1:temp_max) {
temp <- summary(glm(log(CSU_R) ~ CSU_Y,
family=gaussian(link = "identity"),
data=dt_data[dt_data$id_group == i,]
))
dt_data[dt_data$id_group == i, CSU_EAPC:=temp$coefficients[[2]]]
dt_data[dt_data$id_group == i, CSU_ST:=temp$coefficients[[4]]]
}
dt_data$CSU_UP <- 100*(exp(dt_data$CSU_EAPC-(qt((1-CI_level)/2, df = Inf)*dt_data$CSU_ST))-1)
dt_data$CSU_LOW <- 100*(exp(dt_data$CSU_EAPC+(qt((1-CI_level)/2, df = Inf)*dt_data$CSU_ST))-1)
dt_data$CSU_EAPC <- 100*(exp(dt_data$CSU_EAPC)-1)
dt_data<- dt_data[,list( CSU_EAPC=mean(CSU_EAPC), CSU_UP=mean(CSU_UP),CSU_LOW=mean(CSU_LOW)), by=group_by]
setnames(dt_data, "CSU_EAPC", var_eapc)
setnames(dt_data, "CSU_UP", paste(var_eapc, "up", sep="_"))
setnames(dt_data, "CSU_LOW", paste(var_eapc, "low", sep="_"))
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
if (bool_dum_by) {
df_data$CSU_dum_by <- NULL
}
return(df_data)
}
core.csu_ageSpecific <-function(df_data,
var_age="age",
var_cases="cases",
var_py="py",
group_by = NULL,
missing_age = NULL,
db_rate = 100000,
logscale=FALSE,
plot_title=NULL,
legend=csu_trend_legend(),
color_trend = NULL,
CI5_comparison=NULL,
linesize = 0.5,
age_label_list = NULL,
log_point=TRUE,
plot_subtitle=NULL,
plot_caption=NULL,
xtitle = "Age at diagnosis",
ytitle = "Age-specific incidence rate per",
label_group_by = waiver())
{
bool_CI5_comp <- FALSE
CI5_cancer_label <- NULL
bool_dum_by <- FALSE
# manage CI5 import
if (!is.null(CI5_comparison)) {
bool_CI5_comp <- TRUE
data(csu_ci5_mean, envir = e <- new.env())
df_CI5 <- e$csu_ci5_mean
dt_CI5 <- data.table(df_CI5)
if (is.character(CI5_comparison)) {
if (!(CI5_comparison%in% dt_CI5$ci5_cancer_label)) {
stop('CI5_comparison value must be a correct cancer label, see documentation: Help(csu_ci5_mean)')
} else {
dt_CI5 <- dt_CI5[dt_CI5$ci5_cancer_label == CI5_comparison, ]
}
} else {
if (is.numeric(CI5_comparison)) {
if (!(CI5_comparison%in% dt_CI5$ci5_cancer_code)) {
stop('CI5_comparison value must be a correct cancer code, see documentation: Help(csu_ci5_mean)')
} else {
dt_CI5 <- dt_CI5[dt_CI5$ci5_cancer_code == CI5_comparison, ]
}
}
}
CI5_cancer_label <- toString(dt_CI5$ci5_cancer_label[1])
}
# manage group by options
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_age, "CSU_A")
setnames(dt_data, var_cases, "CSU_C")
setnames(dt_data, var_py, "CSU_P")
setnames(dt_data, group_by, "CSU_BY")
##group population (use sum)
dt_data <- dt_data[, list(CSU_C=sum(CSU_C),CSU_P=sum(CSU_P)), by=c("CSU_BY", "CSU_A") ]
##change by to factor
dt_data$CSU_BY <- factor(dt_data$CSU_BY)
##to calcul age group
dt_data[CSU_A %in% missing_age,CSU_A:=NA ]
dt_data[is.na(CSU_A),CSU_P:=0 ]
#parse age
dt_data[, temp_label:=CSU_A]
dt_data[,CSU_A := as.numeric(gsub(".*?(\\d{1,3}).*$", "\\1",CSU_A, perl=TRUE))]
if (max(dt_data$CSU_A,na.rm=TRUE) > 25) {
dt_data[,CSU_A := round((CSU_A/5)+1)]
}
dt_temp_label <- unique(dt_data[, c("CSU_A", "temp_label"), with=FALSE])
dt_data[, temp_label:= NULL]
dt_data$CSU_age_factor <- as.numeric(c(as.factor(dt_data$CSU_A)))
dt_data <- merge(dt_data, dt_data[dt_data$CSU_P != 0,list(nb_age_group = max(CSU_age_factor)), by="CSU_BY"], by="CSU_BY")
for (i in 15:17) {
if (i %in% dt_data$nb_age_group) {
dt_data[nb_age_group == i & CSU_age_factor >= i , CSU_C:=sum(CSU_C), by="CSU_BY"] ##add total_know
dt_data[nb_age_group == i & CSU_age_factor > i & !is.na(CSU_age_factor), CSU_C := 0]
}
}
dt_data <- dt_data[CSU_P!=0]
##calcul rate
dt_data$rate <- dt_data$CSU_C/dt_data$CSU_P *db_rate
##create age label:
if (is.null(age_label_list)) {
##create age dummy: 1 2 3 4 --- 18
##regroup case for population with nb of age group < 18
max_age <- max(dt_data$nb_age_group)
if (max_age > 18) {
stop('The function cannot have more than 18 age-group, see documentation: Help(csu_graph_ageSpecific)')
}
age_label <- c("0-4","5-9","10-14","15-19","20-24","25-39","30-34","35-39","40-44", "45-49","50-54","55-59","60-64","65-69","70-74","75-79","80-84","85+")
age_label <- c(age_label[1:(max_age-1)], paste0((max_age-1)*5,"+"))
} else {
age_label <-age_label_list
max_age <- length(age_label)
}
## to calcul breaks
if (logscale) {
min_tick_value <- min(dt_data[rate != 0,]$rate)
} else {
min_tick_value <- 0
}
tick <- core.csu_tick_generator(max = max(dt_data$rate), min=min_tick_value, logscale = logscale )
tick_space <- tick$tick_list[length(tick$tick_list)] - tick$tick_list[length(tick$tick_list)-1]
temp_top <- ceiling(max(dt_data$rate)/tick_space)*tick_space
temp_expand_y <- max(dt_data$rate)/35
temp_expand_y_up <- max(dt_data$rate)+temp_expand_y
if (temp_expand_y_up > temp_top-(tick_space/2)) {
temp_expand_y_up <- temp_top+temp_expand_y
}
th_legend <- list(theme(legend.position="none"))
if (!bool_dum_by & legend$position == "bottom") {
th_legend <- list(theme(
legend.key = element_rect(fill="transparent"),
legend.position = "bottom",
legend.text = element_text(size = 14),
legend.title = element_text(size = 14),
legend.key.size=unit(1,"cm"),
legend.margin = margin(0, 0, 0, 0)
))
}
if (bool_CI5_comp & is.null(age_label_list)) {
if (max_age < 18) {
dt_CI5[CSU_age_factor >= max_age , CSU_C:=sum(CSU_C)] ##add total_know
dt_CI5[ CSU_age_factor >= max_age , CSU_P:=sum(CSU_P)]
dt_CI5 <- dt_CI5[CSU_age_factor <= max_age]
}
dt_CI5$rate <- dt_CI5$CSU_C/dt_CI5$CSU_P *db_rate
if (logscale) {
dt_CI5[rate == 0 , rate:= NA]
}
}
ylim_inf <- min(c(tick$tick_list, tick$tick_minor_list))
ylim_sup <- max(c(tick$tick_list, tick$tick_minor_list))
##csu_plot
if (logscale) {
base_plot <- ggplot(dt_data[, rate := ifelse(rate==0,NA, rate )], aes(CSU_age_factor, rate))
} else {
base_plot <- ggplot(dt_data, aes(CSU_age_factor, rate))
}
if (bool_CI5_comp) {
pos_y_text = - tick_space
if (temp_top/tick_space > 7) {
pos_y_text = pos_y_text*1.5
}
if(is.null(plot_caption)) {
plot_caption <- paste0("- - - - - - : Mean for ",CI5_cancer_label," cancer in CI5 XII")
}
base_plot <- base_plot +
geom_line(data = dt_CI5,
linewidth = 1,
linetype=2,
colour = "grey50",
show.legend=FALSE)
}
csu_plot <- base_plot+
geom_line(aes(color=CSU_BY), linewidth = 1,na.rm=TRUE)+
guides(color = guide_legend(override.aes = list(size=0.75)))+
labs(title = plot_title,
subtitle = plot_subtitle,
caption = plot_caption)+
scale_x_continuous(name = xtitle,
breaks=seq(1, max_age, 1),
labels = age_label,
minor_breaks = NULL,
expand = c(0.015,0.015)
)
if (logscale){
if (log_point) {
csu_plot <- csu_plot +
geom_point(aes(fill=CSU_BY), size = 3,na.rm=TRUE,shape=21,stroke=0.5,colour="black", show.legend=FALSE)
}
csu_plot <- csu_plot +
scale_y_continuous(name = paste(ytitle, formatC(db_rate, format="d", big.mark=",")),
breaks=tick$tick_list,
minor_breaks = tick$tick_minor_list,
limits=c(ylim_inf,ylim_sup),
labels=core.csu_axes_label,
trans = "log10"
)
} else {
csu_plot <- csu_plot +
coord_cartesian( ylim=c(-temp_expand_y, temp_expand_y_up), expand = TRUE)+
scale_y_continuous(name = paste(ytitle, formatC(db_rate, format="d", big.mark=",")),
breaks=tick$tick_list,
labels=core.csu_axes_label,
expand = c(0,0)
)
}
csu_plot <- csu_plot +
theme(
plot.background= element_blank(),
panel.background = element_blank(),
panel.grid.major= element_line(colour = "grey70"),
panel.grid.minor= element_line(colour = "grey70"),
plot.title = element_text(size=16, margin=margin(0,0,15,0),hjust = 0.5),
plot.subtitle = element_text(size=15, margin=margin(0,0,15,0),hjust = 0.5),
plot.caption = element_text(size=10, margin=margin(15,0,0,0)),
axis.title = element_text(size=14),
axis.title.y = element_text(margin=margin(0,15,0,0)),
axis.title.x = element_text(margin=margin(15,0,0,0)),
plot.margin=margin(20,20,20,20),
axis.text = element_text(size=14, colour = "black"),
axis.text.x = element_text(size=14, angle = 60, hjust = 1),
axis.ticks= element_line(colour = "black", linewidth = linesize),
axis.ticks.length = unit(0.2, "cm"),
axis.line.x = element_line(colour = "black",
linewidth = linesize, linetype = "solid"),
axis.line.y = element_line(colour = "black",
linewidth = linesize, linetype = "solid")
)+
th_legend
if (!is.null(color_trend)) {
csu_plot <- csu_plot +
scale_colour_manual(name=legend$title,
labels = label_group_by,
values= color_trend,
drop = FALSE)
if (logscale) {
csu_plot <- csu_plot +
scale_fill_manual(labels = label_group_by,
values= color_trend,
drop = FALSE)
}
} else {
csu_plot <- csu_plot +
scale_colour_discrete(name=legend$title)
}
if (!bool_dum_by & legend$position=="right") {
csu_plot <- csu_plot +
geom_text(data = dt_data[CSU_age_factor == nb_age_group, ],
aes(label = CSU_BY),
hjust=-0.05)+
theme(plot.margin = unit(c(0.5, legend$right_space_margin, 0.5, 0.5), "lines"))
} else {
csu_plot <- csu_plot +
guides(color = guide_legend(nrow=legend$nrow))
}
dt_data$nb_age_group <- NULL
dt_data$CSU_age_factor <- NULL
if (logscale){
dt_data[, rate := ifelse(is.na(rate),0, rate )]
}
#get back age label
dt_data <- merge(dt_data, dt_temp_label, by=("CSU_A"), all.x=TRUE)
temp = grep("CSU_A", colnames(dt_data))
temp2 = length(colnames(dt_data))
dt_data <- dt_data[,c(1:temp, temp2, (temp+2):temp2-1) , with=FALSE ]
dt_data[, CSU_A := NULL]
setnames(dt_data, "temp_label", "CSU_A")
return(list(csu_plot = csu_plot, dt_data = dt_data, CI5_cancer_label = CI5_cancer_label,legend_position=legend$position,bool_dum_by = bool_dum_by))
}
core.csu_ageSpecific_top <- function(df_data,
var_age,
var_cases,
var_py,
var_top,
group_by=NULL,
missing_age=NULL,
db_rate = 100000,
logscale = FALSE,
nb_top = 5,
plot_title=NULL,
plot_subtitle=NULL,
label_group_by=NULL,
xtitle = "Age at diagnosis",
ytitle = "Age-specific incidence rate per",
var_color=NULL,
plot_caption=NULL,
var_age_label_list = NULL,
caption_bypass=FALSE) {
bool_dum_by <- FALSE
if (is.null(group_by)) {
df_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
df_data$CSU_dum_by <- factor(df_data[[group_by]],levels=c("dummy_by"), labels=c(""))
bool_dum_by <- TRUE
} else {
df_data[[group_by]] <- as.factor(df_data[[group_by]])
}
df_data <- core.csu_dt_rank(df_data, var_value = var_cases, var_rank = var_top,group_by = group_by, number = nb_top)
df_data[["dummy_top"]] <-core.csu_legend_wrapper(df_data[[var_top]], 14)
plotlist <- list()
datalist <- list()
j <- 1
#dummmy variable to factorize variable
df_data$CSU_dum_by <- as.factor(df_data[[group_by]])
for (i in levels( df_data$CSU_dum_by)) {
if (!is.null(label_group_by)) {
label_group <- label_group_by[j]
}
else {
label_group <- i
}
if (caption_bypass) {
if (j == 1) {
plot_caption <- ""
} else {
plot_caption <- plot_title
plot_title <- ""
}
}
dt_plot <- df_data[get("CSU_dum_by") == i]
if (!is.null(var_color)) {
dt_label_order <- setkey(unique(dt_plot[, c("dummy_top",var_color, "CSU_RANK"), with=FALSE]), CSU_RANK)
dt_plot$dummy_top <- factor(dt_plot$dummy_top,levels = dt_label_order$dummy_top)
color_trend <- as.character(dt_label_order[[var_color]])
} else {
dt_label_order <- setkey(unique(dt_plot[, c(var_top, "CSU_RANK"), with=FALSE]), CSU_RANK)
dt_plot[[var_top]] <- factor(dt_plot[[var_top]],levels = dt_label_order[[var_top]])
color_trend <- NULL
}
if (!is.null(var_age_label_list)) {
age_label_list <- unique(dt_plot[[var_age_label_list]])
} else {
age_label_list <- NULL
}
if (is.null(plot_subtitle)) {
subtitle_temp <- label_group
} else {
subtitle_temp <- paste0(plot_subtitle,"\n",label_group)
}
temp <- core.csu_ageSpecific(
dt_plot,
var_age=var_age,
var_cases= var_cases,
var_py=var_py,
group_by = "dummy_top",
missing_age = missing_age,
db_rate= db_rate,
plot_title = plot_title,
plot_subtitle = subtitle_temp,
plot_caption = plot_caption,
color_trend = color_trend,
logscale = logscale,
log_point=FALSE,
age_label_list = age_label_list,
xtitle = xtitle,
ytitle = ytitle
)
dt_temp <- temp$dt_data
dt_temp[[group_by]] <- i
setnames(dt_temp, "CSU_BY", "dummy_top")
plotlist[[j]] <- temp$csu_plot
datalist[[j]] <- dt_temp
j <- j+1
}
df_data <- unique(df_data[,c("dummy_top",group_by, var_top, "CSU_RANK"), with=FALSE])
dt_final <- NULL
for (i in 1:(j-1)) {
dt_final <- rbind(dt_final, datalist[[i]])
}
dt_final <- merge(dt_final,df_data,by=c("dummy_top",group_by),all.x=TRUE, all.y=FALSE)
dt_final[, dummy_top :=NULL]
setnames(dt_final, "CSU_A", var_age)
setnames(dt_final, "CSU_C", var_cases)
setnames(dt_final, "CSU_P", var_py)
setkeyv(dt_final, c(group_by,"CSU_RANK",var_age))
if (bool_dum_by) {
dt_final[, CSU_dum_by:=NULL]
}
return(list(plotlist=plotlist, dt_data=dt_final))
}
core.csu_time_trend <- function (
df_data,
var_trend = "asr",
var_year = "year",
group_by = NULL,
logscale = TRUE,
smoothing = NULL,
legend = csu_trend_legend(),
color_trend = NULL,
ytitle = "Age standardized rate per 100000",
xtitle = "Year",
plot_title = "test",
linesize = 0.5,
plot_subtitle = NULL,
plot_caption = NULL) {
if (!is.null(smoothing)) {
if (smoothing == 0) {
smoothing <- NULL
}
}
dt_data <- data.table(df_data, key = group_by)
setnames(dt_data, var_year, "CSU_Y")
setnames(dt_data, var_trend, "CSU_T")
setnames(dt_data, group_by, "CSU_BY")
bool_dum_by <- FALSE
if (is.null(group_by)) {
dt_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
#change by to factor
dt_data$CSU_BY <- factor(dt_data$CSU_BY)
#smooth with loess fonction
if (!is.null(smoothing))
{
dt_data[,CSU_smooth:= loess( CSU_T ~ CSU_Y, span=smoothing)$fitted, by=CSU_BY]
} else {
dt_data[,CSU_smooth:= CSU_T]
}
dt_data[, max_year:=max(CSU_Y), by=CSU_BY]
# to calcul y axes breaks
if (logscale) {
min_tick_value <- min(dt_data[CSU_smooth > 0,]$CSU_smooth)
} else {
min_tick_value <- 0
}
tick <- core.csu_tick_generator(max = max(dt_data$CSU_smooth), min=min_tick_value, logscale = logscale )
tick_space <- tick$tick_list[length(tick$tick_list)] - tick$tick_list[length(tick$tick_list)-1]
#to calcul year axes break
year_tick <- core.csu_year_tick_generator(min(dt_data$CSU_Y),max(dt_data$CSU_Y))
temp_top <- ceiling(max(dt_data$CSU_smooth)/tick_space)*tick_space
temp_expand_y <- max(dt_data$CSU_smooth)/35
temp_expand_y_up <- max(dt_data$CSU_smooth)+temp_expand_y
if (temp_expand_y_up > temp_top-(tick_space/2)) {
temp_expand_y_up <- temp_top+temp_expand_y
}
th_legend <- list(theme(legend.position="none"))
if (!bool_dum_by & legend$position == "bottom") {
th_legend <- list(theme(
legend.key = element_rect(fill="transparent"),
legend.position = "bottom",
legend.text = element_text(size = 12),
legend.title = element_text(size = 12),
legend.key.size=unit(1,"cm"),
legend.margin = margin(0, 0, 0, 0)
))
}
xlim_inf <- min(c(year_tick$tick_list, year_tick$tick_minor_list))
xlim_sup <- max(c(year_tick$tick_list, year_tick$tick_minor_list))
ylim_inf <- min(c(tick$tick_list, tick$tick_minor_list))
ylim_sup <- max(c(tick$tick_list, tick$tick_minor_list))
#csu_plot
if (logscale) {
base_plot <- ggplot(dt_data[, CSU_smooth := ifelse(CSU_smooth==0,NA, CSU_smooth )], aes(CSU_Y, CSU_smooth))
} else {
base_plot <- ggplot(dt_data, aes(CSU_Y, CSU_smooth))
}
csu_plot <- base_plot+
geom_line(aes(color=CSU_BY), linewidth = 0.75,na.rm=TRUE)+
guides(color = guide_legend(override.aes = list(size=0.75)))+
labs(title = plot_title,
subtitle = plot_subtitle,
caption = plot_caption)+
scale_x_continuous(name = xtitle,
breaks=year_tick$tick_list,
limits=c(xlim_inf,xlim_sup),
minor_breaks = year_tick$tick_minor_list,
expand = c(0.015,0.015)
)
if (logscale){
csu_plot <- csu_plot +
scale_y_continuous(name = ytitle,
breaks=tick$tick_list,
minor_breaks = tick$tick_minor_list,
limits=c(ylim_inf,ylim_sup),
labels=core.csu_axes_label,
trans = "log10"
)
} else {
csu_plot <- csu_plot +
coord_cartesian( ylim=c(-temp_expand_y, temp_expand_y_up), expand = TRUE)+
scale_y_continuous(name = ytitle,
breaks=tick$tick_list,
labels=core.csu_axes_label,
expand = c(0,0)
)
}
if (is.null(color_trend)) {
csu_plot <- csu_plot +scale_colour_discrete(name=legend$title)
}
else {
csu_plot <- csu_plot +scale_colour_manual(name=NULL, values=color_trend)
}
csu_plot <- csu_plot +
theme(
plot.background= element_blank(),
panel.background = element_blank(),
panel.grid.major= element_line(colour = "grey70"),
panel.grid.minor= element_line(colour = "grey70"),
plot.title = element_text(size=16, margin=margin(0,0,15,0),hjust = 0.5),
plot.subtitle = element_text(size=15, margin=margin(0,0,15,0),hjust = 0.5),
plot.caption = element_text(size=10, margin=margin(15,0,0,0)),
axis.title = element_text(size=12),
axis.title.y = element_text(margin=margin(0,15,0,0)),
axis.title.x = element_text(margin=margin(15,0,0,0)),
plot.margin=margin(20,20,20,20),
axis.text = element_text(size=12, colour = "black"),
axis.text.x = element_text(size=12, hjust = 0.5),
axis.ticks= element_line(colour = "black", linewidth = linesize),
axis.ticks.length = unit(0.2, "cm"),
axis.line.x = element_line(colour = "black",
linewidth = linesize, linetype = "solid"),
axis.line.y = element_line(colour = "black",
linewidth = linesize, linetype = "solid")
)+
th_legend
if (!bool_dum_by & legend$position=="right") {
csu_plot <- csu_plot +
geom_text(data = dt_data[CSU_Y == max_year, ],
aes(label = CSU_BY),
hjust=0,
nudge_x=0.5)+
theme(plot.margin = unit(c(0.5, legend$right_space_margin, 0.5, 0.5), "lines"))
} else {
csu_plot <- csu_plot +
guides(color = guide_legend(nrow=legend$nrow))
}
return(list(dt_data=dt_data, csu_plot = csu_plot, legend_position=legend$position,bool_dum_by = bool_dum_by))
}
core.csu_year_extract <- function(year_list) {
temp <- gsub("[^\\d]", "", year_list, perl=TRUE)
if (unique(nchar(temp)) == 4 ) {
return(as.numeric(year_list))
}
else if (unique(nchar(temp)) == 6 ) {
test <- all(grepl("(18|19|20)\\d{2}\\d{2}",temp))
if (test) {
return(as.numeric(gsub("(^\\d{4}).+", "\\1", temp, perl=TRUE)))
}
else {
return(as.numeric(gsub(".+(\\d{4}$)", "\\1", temp, perl=TRUE)))
}
}
else if (unique(nchar(temp)) == 8 ) {
test <- all(grepl("(18|19|20)\\d{2}\\d{4}",temp))
if (test) {
return(as.numeric(gsub("(^\\d{4}).+", "\\1", temp, perl=TRUE)))
}
else {
return(as.numeric(gsub(".+(\\d{4}$)", "\\1", temp, perl=TRUE)))
}
}
}
core.csu_icd_group <- function(icd_list) {
bool_follow <- FALSE
icd_first <- icd_list[1]
icd_long <- icd_first
code_active = as.numeric(sub(".+?(\\d+)", "\\1", icd_first))
for (code in icd_list[-1]) {
code_new = as.numeric(sub(".+?(\\d+)", "\\1", code))
bool_follow <- code_new == code_active + 1
if (bool_follow) {
icd_long <- paste0(icd_first, '-', sprintf("%02d", code_new))
}
else {
icd_long <- paste0(icd_long, ',', as.character(code))
icd_first <- icd_long
}
code_active <- code_new
}
return(icd_long)
}
core.csu_icd_ungroup <- function(icd_group) {
icd_group <- gsub("\\s", "", icd_group)
icd_list <- NULL
ICD_reg <-"([A-Za-z]\\d+)(\\W?)(.+)?"
while (nchar(icd_group)>=3) {
icd_start <- sub(ICD_reg, "\\1", icd_group)
icd_mark <- sub(ICD_reg, "\\2", icd_group)
icd_group <- sub(ICD_reg, "\\3", icd_group)
if (icd_mark == "-") {
letter_start <- sub("([A-Za-z])(\\d+)", "\\1", icd_start)
code_start <- sub("([A-Za-z])(\\d+)", "\\2", icd_start)
code_nchar <- nchar(code_start)
code_start <- as.numeric(code_start)
code_end <- as.numeric(sub("[A-Za-z]?(\\d+)(.+)?", "\\1", icd_group))
for (code in code_start:code_end) {
icd_list <- c(icd_list, paste0(letter_start, sprintf(paste0("%0",code_nchar,"d"), code)))
}
icd_group <- sub("([A-Za-z]?\\d+)(\\W?)(.+)?", "\\3", icd_group)
}
else {
icd_list <- c(icd_list, sub("([A-Za-z]\\d+)", "\\1", icd_start))
}
}
return(icd_list)
}
core.csu_group_cases <- function(df_data, var_age ,group_by=NULL,var_cases = NULL,df_ICD = NULL,var_ICD=NULL,var_year = NULL, all_cancer=FALSE) {
label_by <- NULL
dt_data <- data.table(df_data)
bool_dum_by <- FALSE
if (is.null(group_by)) {
dt_data$CSU_dum_by <- "dummy_by"
group_by <- "CSU_dum_by"
bool_dum_by <- TRUE
}
if (is.null(var_cases)) {
var_cases <- "CSU_C"
dt_data[, CSU_C:= 1]
} else {
setnames(dt_data, var_cases, "CSU_C")
}
dt_data <- dt_data[, unique(c("CSU_C", var_age,group_by,var_ICD,var_year)), with = FALSE]
if (!is.null(var_year)) {
dt_data$year <- core.csu_year_extract(dt_data[[var_year]])
if (var_year != "year") {
dt_data[, (var_year) := NULL]
}
group_by <- c(group_by, "year")
}
if (!is.null(df_ICD)) {
# merge with ICD
dt_ICD <- data.table(df_ICD)
setkeyv(dt_ICD,c("LABEL", "ICD"))
dt_ICD <- unique(dt_ICD)
#ungroup ICD code
dt_table <- data.table()
for (row in 1:nrow(dt_ICD)) {
icd_group <- as.character(dt_ICD[row]$ICD)
temp <- core.csu_icd_ungroup(paste(icd_group, collapse=","))
temp <- data.table(ICD_ungroup = temp, ICD =icd_group )
dt_table <- rbind(dt_table, temp)
}
dt_ICD <- merge(dt_ICD, dt_table, by="ICD")
dt_ICD[, ICD:=NULL]
dt_ICD <- unique(dt_ICD)
dt_ICD_unique <- setDT(dt_ICD)[, .N, keyby=ICD_ungroup][N>1,]
if (nrow(dt_ICD_unique) > 0) {
dt_ICD_unique <- merge(dt_ICD_unique,dt_ICD,by="ICD_ungroup", all.x=TRUE)
dt_ICD_unique <- merge(dt_ICD_unique, df_ICD, by="LABEL")
dt_ICD_unique <- dt_ICD_unique[,c("LABEL", "ICD") , with=FALSE]
cat("-----\n")
print(as.data.frame(dt_ICD_unique))
stop(paste0("There is different label with the same ICD code in the ICD file defined"))
}
setnames(dt_ICD, "ICD_ungroup", "ICD")
dt_ICD[, ICD_group:= sapply(LABEL, function(x) {core.csu_icd_group(as.vector(dt_ICD[LABEL == x, ]$ICD))})]
list_ICD <- dt_ICD$ICD
dt_data$temp <- as.character(dt_data[[var_ICD]])
dt_data[, ICD := list_ICD[match(dt_data$temp, list_ICD)]]
dt_data[!is.na(ICD), temp := NA]
dt_data[, temp:= substr(temp, 1, 3)]
dt_data[!is.na(temp), ICD:=list_ICD[match(dt_data[!is.na(temp)]$temp, list_ICD)]]
dt_data <- dt_data[!is.na(ICD), ]
dt_data <- merge(dt_data, dt_ICD, by=c("ICD"))
dt_data[,c("temp", var_ICD,"ICD") := list(NULL, NULL,NULL)]
dt_data[, LABEL := factor(LABEL)]
dt_data[, ICD_group := factor(ICD_group)]
label_by <- c(label_by,"LABEL")
group_by <- c(group_by, "ICD_group")
dt_ICD <- NULL
}
# create age group
dt_data[, age_group:= ((get(var_age) - (get(var_age) %% 5))/5)+1]
dt_data[, age_group := ifelse(age_group > 18 & age_group <31,18,age_group)]
dt_data[, age_group := ifelse(age_group > 18 ,19,age_group)]
dt_data[, c(var_age) := NULL]
dt_data <- dt_data[,list(CSU_C = sum(CSU_C)),by=eval(colnames(dt_data)[!colnames(dt_data) %in% c("CSU_C")])]
#complete missing age group
if (max(dt_data$age_group) == 19) {
temp <- c(1:19)
}
else {
temp <- c(1:18)
}
dt_CJ = dt_data[, do.call(CJ, c(.SD,list(age_group=temp), unique=TRUE)), .SDcols=group_by]
temp <- copy(colnames(dt_CJ))
# add age_group label
dt_CJ[, temp1 := sprintf("%02d",(age_group-1)*5)]
dt_CJ[, temp2 := (age_group*5)-1]
dt_CJ[, age_group_label := ifelse(temp2 == 89, paste0(temp1,"+"), paste0(temp1,"-", sprintf("%02d", temp2)))]
dt_CJ[age_group == 19 , age_group_label := "Unknown"]
dt_CJ[,c("temp1","temp2") := list(NULL, NULL)]
group_by <- c(group_by, "age_group")
##add ICD group label (but #dad is almost pink)
if (!is.null(df_ICD)) {
dt_temp <- unique(dt_data[, c("ICD_group","LABEL"), with=FALSE])
dt_CJ <- merge(dt_CJ, dt_temp, by="ICD_group", all.x=TRUE)
temp <- c(temp, "LABEL")
}
dt_data <- merge(dt_CJ, dt_data,by=temp, all.x=TRUE)[, CSU_C := ifelse(is.na(CSU_C),0, CSU_C )]
keep_by <- group_by[!group_by %in% c("year", "age_group", "ICD_group")]
setkeyv(dt_data, keep_by)
dt_data[ , group_id := .GRP, by = key(dt_data)]
dt_data[,temp := sum(CSU_C),by=group_id]
dt_data <- dt_data[temp>0,]
dt_data[, c("temp","group_id") := NULL]
setkeyv(dt_data, group_by)
if (var_cases=="CSU_C") {
setnames(dt_data, "CSU_C","cases")
var_cases <- NULL
}
else {
setnames(dt_data, "CSU_C",var_cases)
}
if (bool_dum_by) {
dt_data$CSU_dum_by <- NULL
}
#manage data.table input and output
if (is.data.table(df_data))
{
df_data <- dt_data
} else {
df_data <- data.frame(dt_data)
}
return (df_data)
}
Try the Rcan 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.