AMR/R/aa_helper_functions.R

1237 lines
46 KiB
R
Raw Normal View History

2018-02-21 11:52:31 +01:00
# ==================================================================== #
# TITLE #
# Antimicrobial Resistance (AMR) Data Analysis for R #
2018-02-21 11:52:31 +01:00
# #
2019-01-02 23:24:07 +01:00
# SOURCE #
2020-07-08 14:48:06 +02:00
# https://github.com/msberends/AMR #
2018-02-21 11:52:31 +01:00
# #
# LICENCE #
2020-12-27 00:30:28 +01:00
# (c) 2018-2021 Berends MS, Luz CF et al. #
2020-10-08 11:16:03 +02:00
# Developed at the University of Groningen, the Netherlands, in #
# collaboration with non-profit organisations Certe Medical #
2020-10-26 12:23:03 +01:00
# Diagnostics & Advice, and University Medical Center Groningen. #
2018-02-21 11:52:31 +01:00
# #
2019-01-02 23:24:07 +01:00
# This R package is free software; you can freely use and distribute #
# it for both personal and commercial purposes under the terms of the #
# GNU General Public License version 2.0 (GNU GPL-2), as published by #
# the Free Software Foundation. #
# We created this package for both routine data analysis and academic #
# research and it was publicly released in the hope that it will be #
# useful, but it comes WITHOUT ANY WARRANTY OR LIABILITY. #
2020-10-08 11:16:03 +02:00
# #
# Visit our website for the full manual and a complete tutorial about #
# how to conduct AMR data analysis: https://msberends.github.io/AMR/ #
2018-02-21 11:52:31 +01:00
# ==================================================================== #
# faster implementation of left_join than using merge() by poorman - we use match():
pm_left_join <- function(x, y, by = NULL, suffix = c(".x", ".y")) {
if (is.null(by)) {
by <- intersect(names(x), names(y))[1L]
if (is.na(by)) {
stop_("no common column found for pm_left_join()")
}
2020-09-19 11:54:01 +02:00
pm_join_message(by)
} else if (!is.null(names(by))) {
by <- unname(c(names(by), by))
}
if (length(by) == 1) {
by <- rep(by, 2)
}
2020-10-26 12:23:03 +01:00
int_x <- colnames(x) %in% colnames(y) & colnames(x) != by[1]
int_y <- colnames(y) %in% colnames(x) & colnames(y) != by[2]
colnames(x)[int_x] <- paste0(colnames(x)[int_x], suffix[1L])
colnames(y)[int_y] <- paste0(colnames(y)[int_y], suffix[2L])
2020-10-26 12:23:03 +01:00
merged <- cbind(x,
y[match(x[, by[1], drop = TRUE],
y[, by[2], drop = TRUE]),
colnames(y)[!colnames(y) %in% colnames(x) & !colnames(y) == by[2]],
drop = FALSE])
2020-10-26 12:23:03 +01:00
rownames(merged) <- NULL
merged
}
quick_case_when <- function(...) {
vectors <- list(...)
split <- lapply(vectors, function(x) unlist(strsplit(paste(deparse(x), collapse = ""), "~", fixed = TRUE)))
for (i in seq_len(length(vectors))) {
if (eval(parse(text = split[[i]][1]), envir = parent.frame())) {
return(eval(parse(text = split[[i]][2]), envir = parent.frame()))
}
2020-05-16 13:05:47 +02:00
}
return(NA)
2020-05-16 13:05:47 +02:00
}
# No export, no Rd
addin_insert_in <- function() {
2020-06-17 15:14:37 +02:00
import_fn("insertText", "rstudioapi")(" %in% ")
}
# No export, no Rd
addin_insert_like <- function() {
# we want Shift + Ctrl/Cmd + L to iterate over %like%, %unlike%, %like_case%, and %unlike_case%
getActiveDocumentContext <- import_fn("getActiveDocumentContext", "rstudioapi")
insertText <- import_fn("insertText", "rstudioapi")
modifyRange <- import_fn("modifyRange", "rstudioapi")
document_range <- import_fn("document_range", "rstudioapi")
document_position <- import_fn("document_position", "rstudioapi")
context <- getActiveDocumentContext()
current_row <- context$selection[[1]]$range$end[1]
current_col <- context$selection[[1]]$range$end[2]
current_row_txt <- context$contents[current_row]
if (is.null(current_row) || current_row_txt %unlike% "%(un)?like") {
insertText(" %like% ")
return(invisible())
}
pos_preceded_by <- function(txt) {
if (tryCatch(substr(current_row_txt, current_col - nchar(trimws(txt, which = "right")), current_col) == trimws(txt, which = "right"),
error = function(e) FALSE)) {
return(TRUE)
}
tryCatch(substr(current_row_txt, current_col - nchar(txt), current_col) %like% paste0("^", txt),
error = function(e) FALSE)
}
replace_pos <- function(old, with) {
modifyRange(document_range(document_position(current_row, current_col - nchar(old)),
document_position(current_row, current_col)),
text = with,
id = context$id)
}
if (pos_preceded_by(" %like% ")) {
replace_pos(" %like% ", with = " %unlike% ")
} else if (pos_preceded_by(" %unlike% ")) {
replace_pos(" %unlike% ", with = " %like_case% ")
} else if (pos_preceded_by(" %like_case% ")) {
replace_pos(" %like_case% ", with = " %unlike_case% ")
} else if (pos_preceded_by(" %unlike_case% ")) {
replace_pos(" %unlike_case% ", with = " %like% ")
} else {
insertText(" %like% ")
}
}
2020-02-14 19:54:13 +01:00
check_dataset_integrity <- function() {
2020-07-28 18:39:57 +02:00
# check if user overwrote our data sets in their global environment
data_in_pkg <- data(package = "AMR", envir = asNamespace("AMR"))$results[, "Item", drop = TRUE]
data_in_globalenv <- ls(envir = globalenv())
overwritten <- data_in_pkg[data_in_pkg %in% data_in_globalenv]
2020-09-24 00:30:11 +02:00
# exception for example_isolates
overwritten <- overwritten[overwritten != "example_isolates"]
if (length(overwritten) > 0) {
if (length(overwritten) > 1) {
plural <- c("s are", "", "s")
} else {
plural <- c(" is", "s", "")
}
2021-05-04 12:47:33 +02:00
if (message_not_thrown_before("dataset_overwritten")) {
warning_("The following data set", plural[1],
" overwritten by your global environment and prevent", plural[2],
" the AMR package from working correctly: ",
vector_and(overwritten, quotes = "'"),
".\nPlease rename your object", plural[3], ".", call = FALSE)
}
}
2020-07-28 18:39:57 +02:00
# check if other packages did not overwrite our data sets
valid_microorganisms <- TRUE
valid_antibiotics <- TRUE
tryCatch({
valid_microorganisms <- all(c("mo", "fullname", "kingdom", "phylum",
2020-10-26 12:23:03 +01:00
"class", "order", "family", "genus",
"species", "subspecies", "rank",
2020-05-27 16:37:49 +02:00
"species_id", "source", "ref", "prevalence") %in% colnames(microorganisms),
2020-07-08 14:48:06 +02:00
na.rm = TRUE)
valid_antibiotics <- all(c("ab", "atc", "cid", "name", "group",
"atc_group1", "atc_group2", "abbreviations",
2020-10-26 12:23:03 +01:00
"synonyms", "oral_ddd", "oral_units",
"iv_ddd", "iv_units", "loinc") %in% colnames(antibiotics),
na.rm = TRUE)
}, error = function(e) {
# package not yet loaded
require("AMR")
})
stop_if(!valid_microorganisms | !valid_antibiotics,
"the data set `microorganisms` or `antibiotics` was overwritten in your environment because another package with the same object name(s) was loaded _after_ the AMR package, preventing the AMR package from working correctly. Please load the AMR package last.")
2020-02-14 19:54:13 +01:00
invisible(TRUE)
2019-11-23 12:39:57 +01:00
}
2020-09-24 00:30:11 +02:00
search_type_in_df <- function(x, type, info = TRUE) {
2020-11-17 16:57:41 +01:00
meet_criteria(x, allow_class = "data.frame")
meet_criteria(type, allow_class = "character", has_length = 1)
2019-01-15 12:45:24 +01:00
# try to find columns based on type
found <- NULL
2020-10-26 12:23:03 +01:00
2020-11-17 16:57:41 +01:00
# remove attributes from other packages
2020-05-16 13:05:47 +02:00
x <- as.data.frame(x, stringsAsFactors = FALSE)
colnames_formatted <- tolower(generalise_antibiotic_name(colnames(x)))
2020-10-26 12:23:03 +01:00
2019-01-15 12:45:24 +01:00
# -- mo
if (type == "mo") {
if (any(vapply(FUN.VALUE = logical(1), x, is.mo))) {
# take first <mo> column
found <- colnames(x)[vapply(FUN.VALUE = logical(1), x, is.mo)]
} else if ("mo" %in% colnames_formatted &
suppressWarnings(all(x$mo %in% c(NA, microorganisms$mo)))) {
2019-11-04 12:08:08 +01:00
found <- "mo"
} else if (any(colnames_formatted %like_case% "^(mo|microorganism|organism|bacteria|ba[ck]terie)s?$")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "^(mo|microorganism|organism|bacteria|ba[ck]terie)s?$"])
} else if (any(colnames_formatted %like_case% "^(microorganism|organism|bacteria|ba[ck]terie)")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "^(microorganism|organism|bacteria|ba[ck]terie)"])
} else if (any(colnames_formatted %like_case% "species")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "species"])
2019-01-15 12:45:24 +01:00
}
2020-10-26 12:23:03 +01:00
2019-01-15 12:45:24 +01:00
}
# -- key antibiotics
2021-05-24 15:29:17 +02:00
if (type %in% c("keyantibiotics", "keyantimicrobials")) {
if (any(colnames_formatted %like_case% "^key.*(ab|antibiotics|antimicrobials)")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "^key.*(ab|antibiotics|antimicrobials)"])
2019-01-15 12:45:24 +01:00
}
}
# -- date
if (type == "date") {
if (any(colnames_formatted %like_case% "^(specimen date|specimen_date|spec_date)")) {
2019-03-15 13:57:25 +01:00
# WHONET support
found <- sort(colnames(x)[colnames_formatted %like_case% "^(specimen date|specimen_date|spec_date)"])
if (!any(class(pm_pull(x, found)) %in% c("Date", "POSIXct"))) {
stop(font_red(paste0("Found column '", font_bold(found), "' to be used as input for `col_", type,
"`, but this column contains no valid dates. Transform its values to valid dates first.")),
2019-03-15 13:57:25 +01:00
call. = FALSE)
}
} else if (any(vapply(FUN.VALUE = logical(1), x, function(x) inherits(x, c("Date", "POSIXct"))))) {
# take first <Date> column
found <- colnames(x)[vapply(FUN.VALUE = logical(1), x, function(x) inherits(x, c("Date", "POSIXct")))]
2019-01-15 12:45:24 +01:00
}
}
# -- patient id
if (type == "patient_id") {
crit1 <- colnames_formatted %like_case% "^(patient|patid)"
if (any(crit1)) {
found <- colnames(x)[crit1]
} else {
crit2 <- colnames_formatted %like_case% "(identification |patient|pat.*id)"
if (any(crit2)) {
found <- colnames(x)[crit2]
}
2019-01-29 00:06:50 +01:00
}
}
# -- specimen
if (type == "specimen") {
if (any(colnames_formatted %like_case% "(specimen type|spec_type)")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "(specimen type|spec_type)"])
} else if (any(colnames_formatted %like_case% "^(specimen)")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "^(specimen)"])
2019-01-15 12:45:24 +01:00
}
}
2020-02-20 13:19:23 +01:00
# -- UTI (urinary tract infection)
if (type == "uti") {
if (any(colnames_formatted == "uti")) {
found <- colnames(x)[colnames_formatted == "uti"]
} else if (any(colnames_formatted %like_case% "(urine|urinary)")) {
found <- sort(colnames(x)[colnames_formatted %like_case% "(urine|urinary)"])
2020-02-20 13:19:23 +01:00
}
if (!is.null(found)) {
# this column should contain logicals
2020-02-20 13:19:23 +01:00
if (!is.logical(x[, found, drop = TRUE])) {
message_("Column '", font_bold(found), "' found as input for `col_", type,
2020-10-27 15:56:51 +01:00
"`, but this column does not contain 'logical' values (TRUE/FALSE) and was ignored.",
add_fn = font_red)
2020-02-20 13:19:23 +01:00
found <- NULL
}
}
}
found <- found[1]
2020-09-24 00:30:11 +02:00
if (!is.null(found) & info == TRUE) {
if (message_not_thrown_before(fn = paste0("search_", type))) {
msg <- paste0("Using column '", font_bold(found), "' as input for `col_", type, "`.")
if (type %in% c("keyantibiotics", "keyantimicrobials", "specimen")) {
msg <- paste(msg, "Use", font_bold(paste0("col_", type), "= FALSE"), "to prevent this.")
}
message_(msg)
2019-01-15 12:45:24 +01:00
}
}
found
}
2019-03-26 14:24:03 +01:00
2021-04-16 11:41:05 +02:00
is_valid_regex <- function(x) {
regex_at_all <- tryCatch(vapply(FUN.VALUE = logical(1),
X = strsplit(x, ""),
FUN = function(y) any(y %in% c("$", "(", ")", "*", "+", "-",
".", "?", "[", "]", "^", "{",
"|", "}", "\\"),
2021-04-16 12:02:57 +02:00
na.rm = TRUE),
USE.NAMES = FALSE),
2021-04-16 11:41:05 +02:00
error = function(e) rep(TRUE, length(x)))
regex_valid <- vapply(FUN.VALUE = logical(1),
2021-04-16 12:02:57 +02:00
X = x,
2021-04-16 13:24:59 +02:00
FUN = function(y) !"try-error" %in% class(try(grepl(y, "", perl = TRUE),
2021-04-16 11:41:05 +02:00
silent = TRUE)),
USE.NAMES = FALSE)
regex_at_all & regex_valid
2020-09-24 00:30:11 +02:00
}
stop_ifnot_installed <- function(package) {
2019-06-01 20:40:49 +02:00
# no "utils::installed.packages()" since it requires non-staged install since R 3.6.0
# https://developer.r-project.org/Blog/public/2019/02/14/staged-install/index.html
vapply(FUN.VALUE = character(1), package, function(pkg)
2020-06-17 16:29:10 +02:00
tryCatch(get(".packageName", envir = asNamespace(pkg)),
error = function(e) {
if (pkg == "rstudioapi") {
stop("This function only works in RStudio when using R >= 3.2.", call. = FALSE)
2020-06-17 16:29:10 +02:00
} else if (pkg != "base") {
2020-08-26 16:13:40 +02:00
stop("This requires the '", pkg, "' package.",
2020-06-17 16:29:10 +02:00
"\nTry to install it with: install.packages(\"", pkg, "\")",
2020-06-17 15:14:37 +02:00
call. = FALSE)
}
}))
2019-06-01 20:40:49 +02:00
return(invisible())
2019-03-26 14:24:03 +01:00
}
2019-05-10 16:44:59 +02:00
2021-10-05 09:58:08 +02:00
pkg_is_available <- function(pkg, also_load = TRUE, min_version = NULL) {
2021-05-21 20:20:51 +02:00
if (also_load == TRUE) {
out <- suppressWarnings(require(pkg, character.only = TRUE, warn.conflicts = FALSE))
2021-05-21 20:20:51 +02:00
} else {
out <- requireNamespace(pkg, quietly = TRUE)
}
2021-10-05 09:58:08 +02:00
if (!is.null(min_version)) {
out <- out && packageVersion(pkg) >= min_version
}
2021-05-21 20:20:51 +02:00
isTRUE(out)
}
import_fn <- function(name, pkg, error_on_fail = TRUE) {
if (isTRUE(error_on_fail)) {
stop_ifnot_installed(pkg)
}
2020-07-08 14:48:06 +02:00
tryCatch(
2020-12-17 16:22:25 +01:00
# don't use get() to avoid fetching non-API functions
getExportedValue(name = name, ns = asNamespace(pkg)),
error = function(e) {
if (isTRUE(error_on_fail)) {
2020-12-17 16:22:25 +01:00
stop_("function ", name, "() is not an exported object from package '", pkg,
2020-10-26 12:23:03 +01:00
"'. Please create an issue at https://github.com/msberends/AMR/issues. Many thanks!",
2020-08-26 16:13:40 +02:00
call = FALSE)
} else {
return(NULL)
}
})
2020-06-17 15:14:37 +02:00
}
2020-11-10 16:35:56 +01:00
# this alternative wrapper to the message(), warning() and stop() functions:
2020-10-26 12:23:03 +01:00
# - wraps text to never break lines within words
# - ignores formatted text while wrapping
2020-12-17 16:22:25 +01:00
# - adds indentation dependent on the type of message (such as NOTE)
2020-11-10 16:35:56 +01:00
# - can add additional formatting functions like blue or bold text
word_wrap <- function(...,
add_fn = list(),
as_note = FALSE,
width = 0.95 * getOption("width"),
extra_indent = 0) {
2020-10-26 12:23:03 +01:00
msg <- paste0(c(...), collapse = "")
if (isTRUE(as_note)) {
2021-05-12 18:15:03 +02:00
msg <- paste0(pkg_env$info_icon, " ", gsub("^note:? ?", "", msg, ignore.case = TRUE))
}
2020-11-10 16:35:56 +01:00
2020-12-22 00:51:17 +01:00
if (msg %like% "\n") {
# run word_wraps() over every line here, bind them and return again
return(paste0(vapply(FUN.VALUE = character(1),
trimws(unlist(strsplit(msg, "\n")), which = "right"),
2020-12-22 00:51:17 +01:00
word_wrap,
add_fn = add_fn,
as_note = FALSE,
width = width,
extra_indent = extra_indent),
collapse = "\n"))
}
2021-05-19 22:55:42 +02:00
# correct for operators (will add the space later on)
ops <- "([,./><\\]\\[])"
msg <- gsub(paste0(ops, " ", ops), "\\1\\2", msg, perl = TRUE)
2020-10-26 12:23:03 +01:00
# we need to correct for already applied style, that adds text like "\033[31m\"
msg_stripped <- font_stripstyle(msg)
# where are the spaces now?
msg_stripped_wrapped <- paste0(strwrap(msg_stripped,
simplify = TRUE,
2020-11-10 16:35:56 +01:00
width = width),
collapse = "\n")
msg_stripped_wrapped <- paste0(unlist(strsplit(msg_stripped_wrapped, "(\n|\\*\\|\\*)")),
2020-10-26 12:23:03 +01:00
collapse = "\n")
msg_stripped_spaces <- which(unlist(strsplit(msg_stripped, "")) == " ")
2020-11-10 16:35:56 +01:00
msg_stripped_wrapped_spaces <- which(unlist(strsplit(msg_stripped_wrapped, "")) != "\n")
2020-10-26 12:23:03 +01:00
# so these are the indices of spaces that need to be replaced
replace_spaces <- which(!msg_stripped_spaces %in% msg_stripped_wrapped_spaces)
# put it together
msg <- unlist(strsplit(msg, " "))
msg[replace_spaces] <- paste0(msg[replace_spaces], "\n")
2021-05-19 22:55:42 +02:00
# add space around operators again
msg <- gsub(paste0(ops, ops), "\\1 \\2", msg, perl = TRUE)
2020-10-26 12:23:03 +01:00
msg <- paste0(msg, collapse = " ")
msg <- gsub("\n ", "\n", msg, fixed = TRUE)
2020-11-10 16:35:56 +01:00
if (msg_stripped %like% "\u2139 ") {
indentation <- 2 + extra_indent
2020-10-27 15:56:51 +01:00
} else if (msg_stripped %like% "^=> ") {
2020-11-10 16:35:56 +01:00
indentation <- 3 + extra_indent
2020-10-26 12:23:03 +01:00
} else {
2020-11-10 16:35:56 +01:00
indentation <- 0 + extra_indent
2020-10-26 12:23:03 +01:00
}
msg <- gsub("\n", paste0("\n", strrep(" ", indentation)), msg, fixed = TRUE)
# remove trailing empty characters
msg <- gsub("(\n| )+$", "", msg)
2021-05-19 22:55:42 +02:00
2020-10-26 12:23:03 +01:00
if (length(add_fn) > 0) {
if (!is.list(add_fn)) {
add_fn <- list(add_fn)
}
for (i in seq_len(length(add_fn))) {
msg <- add_fn[[i]](msg)
}
}
2020-11-10 16:35:56 +01:00
# format backticks
msg <- gsub("(`.+?`)", font_grey_bg("\\1"), msg)
# clean introduced whitespace between fullstops
msg <- gsub("[.] +[.]", "..", msg)
2020-11-10 16:35:56 +01:00
msg
}
message_ <- function(...,
appendLF = TRUE,
add_fn = list(font_blue),
as_note = TRUE) {
message(word_wrap(...,
add_fn = add_fn,
as_note = as_note),
appendLF = appendLF)
}
warning_ <- function(...,
add_fn = list(),
immediate = FALSE,
call = TRUE) {
warning(word_wrap(...,
add_fn = add_fn,
as_note = FALSE),
immediate. = immediate,
call. = call)
2020-10-26 12:23:03 +01:00
}
# this alternative to the stop() function:
# - adds the function name where the error was thrown
# - wraps text to never break lines within words
stop_ <- function(..., call = TRUE) {
2020-11-28 22:15:44 +01:00
msg <- paste0(c(...), collapse = "")
if (!isFALSE(call)) {
if (isTRUE(call)) {
call <- as.character(sys.call(-1)[1])
} else {
# so you can go back more than 1 call, as used in rsi_calc(), that now throws a reference to e.g. n_rsi()
call <- as.character(sys.call(call)[1])
}
msg <- paste0("in ", call, "(): ", msg)
}
2020-11-28 22:15:44 +01:00
msg <- word_wrap(msg, add_fn = list(), as_note = FALSE)
stop(msg, call. = FALSE)
}
stop_if <- function(expr, ..., call = TRUE) {
if (isTRUE(expr)) {
if (isTRUE(call)) {
call <- -1
}
2020-06-26 10:21:22 +02:00
if (!isFALSE(call)) {
# since we're calling stop_(), which is another call
call <- call - 1
2020-06-26 10:21:22 +02:00
}
stop_(..., call = call)
2020-03-08 11:18:59 +01:00
}
}
stop_ifnot <- function(expr, ..., call = TRUE) {
2020-09-24 00:30:11 +02:00
if (isFALSE(expr)) {
if (isTRUE(call)) {
call <- -1
}
2020-06-26 10:21:22 +02:00
if (!isFALSE(call)) {
# since we're calling stop_(), which is another call
call <- call - 1
2020-06-26 10:21:22 +02:00
}
stop_(..., call = call)
}
}
2020-03-08 11:18:59 +01:00
2019-05-20 19:12:41 +02:00
"%or%" <- function(x, y) {
if (is.null(x) | is.null(y)) {
if (is.null(x)) {
return(y)
} else {
return(x)
}
}
ifelse(!is.na(x),
x,
ifelse(!is.na(y), y, NA))
2019-05-20 19:12:41 +02:00
}
return_after_integrity_check <- function(value, type, check_vector) {
if (!all(value[!is.na(value)] %in% check_vector)) {
2020-11-10 16:35:56 +01:00
warning_(paste0("invalid ", type, ", NA generated"), call = FALSE)
value[!value %in% check_vector] <- NA
}
value
}
2020-01-27 19:14:23 +01:00
# transforms data set to data.frame with only ASCII values, to comply with CRAN policies
dataset_UTF8_to_ASCII <- function(df) {
trans <- function(vect) {
iconv(vect, from = "UTF-8", to = "ASCII//TRANSLIT")
}
df <- as.data.frame(df, stringsAsFactors = FALSE)
for (i in seq_len(NCOL(df))) {
col <- df[, i]
if (is.list(col)) {
col <- lapply(col, function(j) trans(j))
df[, i] <- list(col)
} else {
if (is.factor(col)) {
levels(col) <- trans(levels(col))
} else if (is.character(col)) {
col <- trans(col)
} else {
col
}
df[, i] <- col
}
}
df
}
2020-05-16 13:05:47 +02:00
2020-11-28 22:15:44 +01:00
# for eucast_rules() and mdro(), creates markdown output with URLs and names
2021-05-18 11:29:31 +02:00
create_eucast_ab_documentation <- function() {
x <- trimws(unique(toupper(unlist(strsplit(EUCAST_RULES_DF$then_change_these_antibiotics, ",")))))
2021-05-18 11:29:31 +02:00
ab <- character()
for (val in x) {
2021-08-17 14:34:11 +02:00
if (paste0("AB_", val) %in% ls(envir = asNamespace("AMR"))) {
2021-05-18 11:29:31 +02:00
# antibiotic group names, as defined in data-raw/_internals.R, such as `CARBAPENEMS`
2021-08-17 14:34:11 +02:00
val <- eval(parse(text = paste0("AB_", val)), envir = asNamespace("AMR"))
2021-05-18 11:29:31 +02:00
} else if (val %in% AB_lookup$ab) {
# separate drugs, such as `AMX`
val <- as.ab(val)
} else {
2021-05-20 00:07:27 +02:00
val <- as.rsi(NA)
2021-05-18 11:29:31 +02:00
}
ab <- c(ab, val)
}
ab <- unique(ab)
2021-08-17 14:34:11 +02:00
atcs <- ab_atc(ab, only_first = TRUE)
2021-05-18 11:29:31 +02:00
# only keep ABx with an ATC code:
ab <- ab[!is.na(atcs)]
2020-09-24 00:30:11 +02:00
ab_names <- ab_name(ab, language = NULL, tolower = TRUE)
ab <- ab[order(ab_names)]
ab_names <- ab_names[order(ab_names)]
2021-05-18 11:29:31 +02:00
atc_txt <- paste0("[", atcs[!is.na(atcs)], "](", ab_url(ab), ")")
out <- paste0(ab_names, " (`", ab, "`, ", atc_txt, ")", collapse = ", ")
2020-09-24 00:30:11 +02:00
substr(out, 1, 1) <- toupper(substr(out, 1, 1))
out
}
vector_or <- function(v, quotes = TRUE, reverse = FALSE, sort = TRUE, initial_captital = FALSE, last_sep = " or ") {
# makes unique and sorts, and this also removed NAs
v <- unique(v)
if (isTRUE(sort)) {
v <- sort(v)
}
if (isTRUE(reverse)) {
v <- rev(v)
}
if (isTRUE(quotes)) {
quotes <- '"'
} else if (isFALSE(quotes)) {
quotes <- ""
} else {
quotes <- quotes[1L]
}
if (isTRUE(initial_captital)) {
v[1] <- gsub("^([a-z])", "\\U\\1", v[1], perl = TRUE)
}
if (length(v) == 1) {
return(paste0(quotes, v, quotes))
}
if (identical(v, c("I", "R", "S"))) {
# class <rsi> should be sorted like this
v <- c("R", "S", "I")
}
# all commas except for last item, so will become '"val1", "val2", "val3" or "val4"'
paste0(paste0(quotes, v[seq_len(length(v) - 1)], quotes, collapse = ", "),
last_sep, paste0(quotes, v[length(v)], quotes))
}
vector_and <- function(v, quotes = TRUE, reverse = FALSE, sort = TRUE, initial_captital = FALSE) {
vector_or(v = v, quotes = quotes, reverse = reverse, sort = sort,
initial_captital = initial_captital, last_sep = " and ")
}
2021-05-12 18:15:03 +02:00
format_class <- function(class, plural = FALSE) {
class.bak <- class
class[class == "numeric"] <- "number"
class[class == "integer"] <- "whole number"
if (all(c("numeric", "integer") %in% class.bak, na.rm = TRUE)) {
class[class %in% c("number", "whole number")] <- "(whole) number"
}
class[class == "character"] <- "text string"
class[class %in% c("Date", "POSIXt")] <- "date"
class[class != class.bak] <- paste0(ifelse(plural, "", "a "),
class[class != class.bak],
ifelse(plural, "s", ""))
# exceptions
class[class == "logical"] <- ifelse(plural, "a vector of `TRUE`/`FALSE`", "`TRUE` or `FALSE`")
2021-05-12 18:15:03 +02:00
class[class == "data.frame"] <- "a data set"
if ("list" %in% class) {
class <- "a list"
}
if ("matrix" %in% class) {
class <- "a matrix"
}
2021-04-07 08:37:42 +02:00
if ("custom_eucast_rules" %in% class) {
class <- "input created with `custom_eucast_rules()`"
2021-01-22 10:55:07 +01:00
}
if (any(c("mo", "ab", "rsi") %in% class)) {
2021-01-22 10:55:07 +01:00
class <- paste0("of class <", class[1L], ">")
}
2021-01-22 10:55:07 +01:00
class[class == class.bak] <- paste0("of class <", class[class == class.bak], ">")
# output
vector_or(class, quotes = FALSE, sort = FALSE)
}
# a check for every single argument in all functions
meet_criteria <- function(object,
allow_class = NULL,
has_length = NULL,
looks_like = NULL,
is_in = NULL,
is_positive = NULL,
2021-04-07 08:37:42 +02:00
is_positive_or_zero = NULL,
is_finite = NULL,
contains_column_class = NULL,
allow_NULL = FALSE,
allow_NA = FALSE,
ignore.case = FALSE,
.call_depth = 0) { # depth in calling
2020-10-26 12:23:03 +01:00
obj_name <- deparse(substitute(object))
call_depth <- -2 - abs(.call_depth)
# if object is missing, or another error:
tryCatch(invisible(object),
error = function(e) pkg_env$meet_criteria_error_txt <- e$message)
if (!is.null(pkg_env$meet_criteria_error_txt)) {
error_txt <- pkg_env$meet_criteria_error_txt
pkg_env$meet_criteria_error_txt <- NULL
stop(error_txt, call. = FALSE) # don't use stop_() here, pkg may not be loaded yet
}
pkg_env$meet_criteria_error_txt <- NULL
2020-10-26 12:23:03 +01:00
if (is.null(object)) {
stop_if(allow_NULL == FALSE, "argument `", obj_name, "` must not be NULL", call = call_depth)
return(invisible())
}
2021-01-04 12:29:25 +01:00
if (is.null(dim(object)) && length(object) == 1 && suppressWarnings(is.na(object))) { # suppressWarnings for functions
stop_if(allow_NA == FALSE, "argument `", obj_name, "` must not be NA", call = call_depth)
return(invisible())
}
2020-10-26 12:23:03 +01:00
if (!is.null(allow_class)) {
2020-10-26 12:23:03 +01:00
stop_ifnot(inherits(object, allow_class), "argument `", obj_name,
"` must be ", format_class(allow_class, plural = isTRUE(has_length > 1)),
2021-01-22 10:55:07 +01:00
", i.e. not be ", format_class(class(object), plural = isTRUE(has_length > 1)),
call = call_depth)
# check data.frames for data
if (inherits(object, "data.frame")) {
2020-10-26 12:23:03 +01:00
stop_if(any(dim(object) == 0),
"the data provided in argument `", obj_name,
2020-10-26 12:23:03 +01:00
"` must contain rows and columns (current dimensions: ",
paste(dim(object), collapse = "x"), ")",
call = call_depth)
}
}
if (!is.null(has_length)) {
2020-10-26 12:23:03 +01:00
stop_ifnot(length(object) %in% has_length, "argument `", obj_name,
"` must ", # ifelse(allow_NULL, "be NULL or must ", ""),
"be of length ", vector_or(has_length, quotes = FALSE),
", not ", length(object),
call = call_depth)
}
if (!is.null(looks_like)) {
2020-10-26 12:23:03 +01:00
stop_ifnot(object %like% looks_like, "argument `", obj_name,
"` must ", # ifelse(allow_NULL, "be NULL or must ", ""),
"resemble the regular expression \"", looks_like, "\"",
call = call_depth)
}
if (!is.null(is_in)) {
if (ignore.case == TRUE) {
object <- tolower(object)
is_in <- tolower(is_in)
}
2021-05-18 11:29:31 +02:00
stop_ifnot(all(object %in% is_in, na.rm = TRUE), "argument `", obj_name, "` ",
ifelse(!is.null(has_length) && length(has_length) == 1 && has_length == 1,
"must be either ",
"must only contain values "),
vector_or(is_in, quotes = !isTRUE(any(c("double", "numeric", "integer") %in% allow_class))),
ifelse(allow_NA == TRUE, ", or NA", ""),
call = call_depth)
}
2021-04-12 12:35:13 +02:00
if (isTRUE(is_positive)) {
stop_if(is.numeric(object) && !all(object > 0, na.rm = TRUE), "argument `", obj_name,
"` must ",
ifelse(!is.null(has_length) && length(has_length) == 1 && has_length == 1,
2021-04-07 08:37:42 +02:00
"be a number higher than zero",
"all be numbers higher than zero"),
call = call_depth)
}
2021-04-12 12:35:13 +02:00
if (isTRUE(is_positive_or_zero)) {
2021-04-07 08:37:42 +02:00
stop_if(is.numeric(object) && !all(object >= 0, na.rm = TRUE), "argument `", obj_name,
"` must ",
ifelse(!is.null(has_length) && length(has_length) == 1 && has_length == 1,
"be zero or a positive number",
"all be zero or numbers higher than zero"),
call = call_depth)
}
2021-04-12 12:35:13 +02:00
if (isTRUE(is_finite)) {
stop_if(is.numeric(object) && !all(is.finite(object[!is.na(object)]), na.rm = TRUE), "argument `", obj_name,
"` must ",
ifelse(!is.null(has_length) && length(has_length) == 1 && has_length == 1,
"be a finite number",
"all be finite numbers"),
" (i.e. not be infinite)",
call = call_depth)
}
if (!is.null(contains_column_class)) {
stop_ifnot(any(vapply(FUN.VALUE = logical(1),
object,
function(col, columns_class = contains_column_class) {
inherits(col, columns_class)
}), na.rm = TRUE),
2020-10-26 12:23:03 +01:00
"the data provided in argument `", obj_name,
"` must contain at least one column of class <", contains_column_class, ">. ",
"See ?as.", contains_column_class, ".",
call = call_depth)
}
return(invisible())
}
get_current_data <- function(arg_name, call) {
# try dplyr::cur_data_all() first to support dplyr groups
# only useful for e.g. dplyr::filter(), dplyr::mutate() and dplyr::summarise()
# not useful (throws error) with e.g. dplyr::select() - but that will be caught later in this function
cur_data_all <- import_fn("cur_data_all", "dplyr", error_on_fail = FALSE)
if (!is.null(cur_data_all)) {
out <- tryCatch(cur_data_all(), error = function(e) NULL)
if (is.data.frame(out)) {
return(structure(out, type = "dplyr_cur_data_all"))
2021-01-04 12:29:25 +01:00
}
}
# try a manual (base R) method, by going over all underlying environments with sys.frames()
for (env in sys.frames()) {
if (!is.null(env$`.Generic`)) {
# don't check `".Generic" %in% names(env)`, because in R < 3.2, `names(env)` is always NULL
if (!is.null(env$`.data`) && is.data.frame(env$`.data`)) {
# an element `.data` will be in the environment when using `dplyr::select()`
# (but not when using `dplyr::filter()`, `dplyr::mutate()` or `dplyr::summarise()`)
return(structure(env$`.data`, type = "dplyr_selector"))
} else if (!is.null(env$xx) && is.data.frame(env$xx)) {
# an element `xx` will be in the environment for rows + cols, e.g. `example_isolates[c(1:3), carbapenems()]`
return(structure(env$xx, type = "base_R"))
} else if (!is.null(env$x) && is.data.frame(env$x)) {
# an element `x` will be in the environment for only cols, e.g. `example_isolates[, carbapenems()]`
return(structure(env$x, type = "base_R"))
}
}
}
# no data.frame found, so an error must be returned:
if (is.na(arg_name)) {
if (isTRUE(is.numeric(call))) {
fn <- as.character(sys.call(call + 1)[1])
examples <- paste0(", e.g.:\n",
" your_data %>% select(", fn, "())\n",
" your_data %>% select(column_a, column_b, ", fn, "())\n",
" your_data[, ", fn, "()]\n",
' your_data[, c("column_a", "column_b", ', fn, "())]")
} else {
examples <- ""
}
stop_("this function must be used inside a `dplyr` verb or `data.frame` call",
examples,
call = call)
} else {
# mimic a base R error that the argument is missing
stop_("argument `", arg_name, "` is missing with no default", call = call)
}
}
get_current_column <- function() {
# try dplyr::cur_columns() first
cur_column <- import_fn("cur_column", "dplyr", error_on_fail = FALSE)
if (!is.null(cur_column)) {
out <- tryCatch(cur_column(), error = function(e) NULL)
if (!is.null(out)) {
return(out)
}
}
# cur_column() doesn't always work (only allowed for certain conditions set by dplyr), but it's probably still possible:
frms <- lapply(sys.frames(), function(env) {
if (!is.null(env$i)) {
if (!is.null(env$tibble_vars)) {
# for mutate_if()
env$tibble_vars[env$i]
} else {
# for mutate(across())
df <- tryCatch(get_current_data(NA, 0), error = function(e) NULL)
if (is.data.frame(df)) {
colnames(df)[env$i]
} else {
env$i
}
}
} else {
NULL
}
})
vars <- unlist(frms)
if (length(vars) > 0) {
vars[length(vars)]
} else {
# not found, so:
NULL
}
}
is_null_or_grouped_tbl <- function(x) {
# class "grouped_df" might change at one point, so only set in one place; here.
2021-02-08 21:09:36 +01:00
is.null(x) || inherits(x, "grouped_df")
}
unique_call_id <- function(entire_session = FALSE) {
if (entire_session == TRUE) {
c(envir = "session",
call = "session")
} else {
# combination of environment ID (such as "0x7fed4ee8c848")
# and highest system call
2021-05-24 11:01:32 +02:00
call <- paste0(deparse(sys.calls()[[1]]), collapse = "")
2021-05-24 15:29:17 +02:00
if (!interactive() || call %like% "run_test_dir|test_all|tinytest|test_package|testthat") {
2021-05-24 11:01:32 +02:00
# unit tests will keep the same call and environment - give them a unique ID
call <- paste0(sample(c(c(0:9), letters[1:6]), size = 64, replace = TRUE), collapse = "")
}
c(envir = gsub("<environment: (.*)>", "\\1", utils::capture.output(sys.frames()[[1]])),
2021-05-24 11:01:32 +02:00
call = call)
}
}
message_not_thrown_before <- function(fn, entire_session = FALSE) {
# this is to prevent that messages/notes will be printed for every dplyr group or more than once per session
2020-12-27 00:07:00 +01:00
# e.g. this would show a msg 4 times: example_isolates %>% group_by(hospital_id) %>% filter(mo_is_gram_negative())
not_thrown_before <- is.null(pkg_env[[paste0("thrown_msg.", fn)]]) || !identical(pkg_env[[paste0("thrown_msg.", fn)]],
unique_call_id(entire_session = entire_session))
if (isTRUE(not_thrown_before)) {
# message was not thrown before - remember this so on the next run it will return FALSE:
assign(x = paste0("thrown_msg.", fn),
value = unique_call_id(entire_session = entire_session),
envir = pkg_env)
}
not_thrown_before
2020-12-27 00:07:00 +01:00
}
2020-06-17 15:14:37 +02:00
has_colour <- function() {
2020-12-27 00:07:00 +01:00
# this is a base R version of crayon::has_color, but disables colours on emacs
if (Sys.getenv("EMACS") != "" || Sys.getenv("INSIDE_EMACS") != "") {
# disable on emacs, which only supports 8 colours
2020-12-27 00:07:00 +01:00
return(FALSE)
}
enabled <- getOption("crayon.enabled")
if (!is.null(enabled)) {
return(isTRUE(enabled))
}
rstudio_with_ansi_support <- function(x) {
if (Sys.getenv("RSTUDIO", "") == "") {
return(FALSE)
}
2021-05-24 00:06:28 +02:00
if ((cols <- Sys.getenv("RSTUDIO_CONSOLE_COLOR", "")) != "" && !is.na(as.double(cols))) {
return(TRUE)
}
tryCatch(get("isAvailable", envir = asNamespace("rstudioapi"))(), error = function(e) return(FALSE)) &&
tryCatch(get("hasFun", envir = asNamespace("rstudioapi"))("getConsoleHasColor"), error = function(e) return(FALSE))
}
if (rstudio_with_ansi_support() && sink.number() == 0) {
return(TRUE)
}
if (!isatty(stdout())) {
2020-05-16 13:05:47 +02:00
return(FALSE)
}
if (tolower(Sys.info()["sysname"]) == "windows") {
if (Sys.getenv("ConEmuANSI") == "ON") {
2020-05-16 13:05:47 +02:00
return(TRUE)
}
if (Sys.getenv("CMDER_ROOT") != "") {
return(TRUE)
}
return(FALSE)
}
if ("COLORTERM" %in% names(Sys.getenv())) {
return(TRUE)
}
if (Sys.getenv("TERM") == "dumb") {
return(FALSE)
}
2020-10-26 12:23:03 +01:00
grepl(pattern = "^screen|^xterm|^vt100|color|ansi|cygwin|linux",
x = Sys.getenv("TERM"),
ignore.case = TRUE,
perl = TRUE)
2020-05-16 13:05:47 +02:00
}
# set colours if console has_colour()
2020-05-16 13:05:47 +02:00
try_colour <- function(..., before, after, collapse = " ") {
txt <- paste0(unlist(list(...)), collapse = collapse)
if (isTRUE(has_colour())) {
if (is.null(collapse)) {
paste0(before, txt, after, collapse = NULL)
} else {
paste0(before, txt, after, collapse = "")
}
} else {
txt
}
}
font_black <- function(..., collapse = " ") {
try_colour(..., before = "\033[38;5;232m", after = "\033[39m", collapse = collapse)
}
font_blue <- function(..., collapse = " ") {
try_colour(..., before = "\033[34m", after = "\033[39m", collapse = collapse)
}
font_green <- function(..., collapse = " ") {
try_colour(..., before = "\033[32m", after = "\033[39m", collapse = collapse)
}
font_magenta <- function(..., collapse = " ") {
try_colour(..., before = "\033[35m", after = "\033[39m", collapse = collapse)
}
font_red <- function(..., collapse = " ") {
try_colour(..., before = "\033[31m", after = "\033[39m", collapse = collapse)
}
font_silver <- function(..., collapse = " ") {
try_colour(..., before = "\033[90m", after = "\033[39m", collapse = collapse)
}
font_white <- function(..., collapse = " ") {
try_colour(..., before = "\033[37m", after = "\033[39m", collapse = collapse)
}
font_yellow <- function(..., collapse = " ") {
try_colour(..., before = "\033[33m", after = "\033[39m", collapse = collapse)
}
font_subtle <- function(..., collapse = " ") {
try_colour(..., before = "\033[38;5;246m", after = "\033[39m", collapse = collapse)
}
font_grey <- function(..., collapse = " ") {
try_colour(..., before = "\033[38;5;249m", after = "\033[39m", collapse = collapse)
}
font_grey_bg <- function(..., collapse = " ") {
2021-08-17 14:34:11 +02:00
if (tryCatch(import_fn("getThemeInfo", "rstudioapi", error_on_fail = FALSE)()$dark, error = function(e) FALSE)) {
# similar to HTML #444444
try_colour(..., before = "\033[48;5;238m", after = "\033[49m", collapse = collapse)
} else {
# similar to HTML #f0f0f0
try_colour(..., before = "\033[48;5;255m", after = "\033[49m", collapse = collapse)
}
}
2020-05-16 13:05:47 +02:00
font_green_bg <- function(..., collapse = " ") {
try_colour(..., before = "\033[42m", after = "\033[49m", collapse = collapse)
}
2020-12-27 00:07:00 +01:00
font_rsi_R_bg <- function(..., collapse = " ") {
#ED553B
try_colour(..., before = "\033[48;5;203m", after = "\033[49m", collapse = collapse)
2020-12-27 00:07:00 +01:00
}
font_rsi_S_bg <- function(..., collapse = " ") {
#3CAEA3
try_colour(..., before = "\033[48;5;79m", after = "\033[49m", collapse = collapse)
2020-12-27 00:07:00 +01:00
}
font_rsi_I_bg <- function(..., collapse = " ") {
#F6D55C
try_colour(..., before = "\033[48;5;222m", after = "\033[49m", collapse = collapse)
2020-12-27 00:07:00 +01:00
}
2020-05-16 13:05:47 +02:00
font_red_bg <- function(..., collapse = " ") {
try_colour(..., before = "\033[41m", after = "\033[49m", collapse = collapse)
}
font_yellow_bg <- function(..., collapse = " ") {
try_colour(..., before = "\033[43m", after = "\033[49m", collapse = collapse)
}
font_na <- function(..., collapse = " ") {
font_red(..., collapse = collapse)
}
2020-05-16 13:05:47 +02:00
font_bold <- function(..., collapse = " ") {
try_colour(..., before = "\033[1m", after = "\033[22m", collapse = collapse)
}
font_italic <- function(..., collapse = " ") {
try_colour(..., before = "\033[3m", after = "\033[23m", collapse = collapse)
}
font_underline <- function(..., collapse = " ") {
try_colour(..., before = "\033[4m", after = "\033[24m", collapse = collapse)
}
font_stripstyle <- function(x) {
# from crayon:::ansi_regex
gsub("(?:(?:\\x{001b}\\[)|\\x{009b})(?:(?:[0-9]{1,3})?(?:(?:;[0-9]{0,3})*)?[A-M|f-m])|\\x{001b}[A-M]", "", x, perl = TRUE)
}
progress_ticker <- function(n = 1, n_min = 0, print = TRUE, ...) {
if (print == FALSE || n < n_min) {
2020-05-16 13:05:47 +02:00
pb <- list()
pb$tick <- function() {
invisible()
}
pb$kill <- function() {
invisible()
}
set_clean_class(pb, new_class = "txtProgressBar")
2020-07-29 10:33:47 +02:00
} else if (n >= n_min) {
pb <- utils::txtProgressBar(max = n, style = 3)
pb$tick <- function() {
pb$up(pb$getVal() + 1)
}
pb
2020-05-16 13:05:47 +02:00
}
}
set_clean_class <- function(x, new_class) {
# return the object with only the new class and no additional attributes where possible
if (is.null(x)) {
x <- NA_character_
}
if (is.factor(x)) {
# keep only levels and remove all other attributes
lvls <- levels(x)
attributes(x) <- NULL
levels(x) <- lvls
2020-11-16 20:02:20 +01:00
} else if (!is.list(x) && !is.function(x)) {
attributes(x) <- NULL
}
class(x) <- new_class
x
}
2020-12-17 16:22:25 +01:00
formatted_filesize <- function(...) {
size_kb <- file.size(...) / 1024
if (size_kb < 1) {
paste(round(size_kb, 1), "kB")
} else if (size_kb < 100) {
paste(round(size_kb, 0), "kB")
} else {
paste(round(size_kb / 1024, 1), "MB")
}
}
create_pillar_column <- function(x, ...) {
new_pillar_shaft_simple <- import_fn("new_pillar_shaft_simple", "pillar")
new_pillar_shaft_simple(x, ...)
}
# copied from vctrs::s3_register by their permission:
# https://github.com/r-lib/vctrs/blob/05968ce8e669f73213e3e894b5f4424af4f46316/R/register-s3.R
s3_register <- function(generic, class, method = NULL) {
stopifnot(is.character(generic), length(generic) == 1)
stopifnot(is.character(class), length(class) == 1)
pieces <- strsplit(generic, "::")[[1]]
stopifnot(length(pieces) == 2)
package <- pieces[[1]]
generic <- pieces[[2]]
caller <- parent.frame()
get_method_env <- function() {
top <- topenv(caller)
if (isNamespace(top)) {
asNamespace(environmentName(top))
}
else {
caller
}
}
get_method <- function(method, env) {
if (is.null(method)) {
get(paste0(generic, ".", class), envir = get_method_env())
}
else {
method
}
}
method_fn <- get_method(method)
stopifnot(is.function(method_fn))
setHook(packageEvent(package, "onLoad"), function(...) {
ns <- asNamespace(package)
method_fn <- get_method(method)
registerS3method(generic, class, method_fn, envir = ns)
})
if (!isNamespaceLoaded(package)) {
return(invisible())
}
envir <- asNamespace(package)
if (exists(generic, envir)) {
registerS3method(generic, class, method_fn, envir = envir)
}
invisible()
}
# works exactly like round(), but rounds `round2(44.55, 1)` to 44.6 instead of 44.5
2020-05-16 13:05:47 +02:00
# and adds decimal zeroes until `digits` is reached when force_zero = TRUE
2021-05-24 15:29:17 +02:00
round2 <- function(x, digits = 1, force_zero = TRUE) {
2020-05-16 13:05:47 +02:00
x <- as.double(x)
# https://stackoverflow.com/a/12688836/4575331
val <- (trunc((abs(x) * 10 ^ digits) + 0.5) / 10 ^ digits) * sign(x)
if (digits > 0 & force_zero == TRUE) {
values_trans <- val[val != as.integer(val) & !is.na(val)]
val[val != as.integer(val) & !is.na(val)] <- paste0(values_trans,
2020-10-26 12:23:03 +01:00
strrep("0",
max(0,
2020-05-16 13:05:47 +02:00
digits - nchar(
format(
as.double(
2020-10-26 12:23:03 +01:00
gsub(".*[.](.*)$",
2020-05-16 13:05:47 +02:00
"\\1",
values_trans)),
scientific = FALSE)))))
}
as.double(val)
}
# percentage from our other package: 'cleaner'
percentage <- function(x, digits = NULL, ...) {
2020-10-26 12:23:03 +01:00
2020-05-16 13:05:47 +02:00
# getdecimalplaces() function
getdecimalplaces <- function(x, minimum = 0, maximum = 3) {
if (maximum < minimum) {
maximum <- minimum
}
if (minimum > maximum) {
minimum <- maximum
}
2020-10-26 12:23:03 +01:00
max_places <- max(unlist(lapply(strsplit(sub("0+$", "",
2020-05-16 13:05:47 +02:00
as.character(x * 100)), ".", fixed = TRUE),
function(y) ifelse(length(y) == 2, nchar(y[2]), 0))), na.rm = TRUE)
max(min(max_places,
maximum, na.rm = TRUE),
minimum, na.rm = TRUE)
}
2020-10-26 12:23:03 +01:00
2020-05-16 13:05:47 +02:00
# format_percentage() function
format_percentage <- function(x, digits = NULL, ...) {
if (is.null(digits)) {
digits <- getdecimalplaces(x)
}
2021-05-20 11:42:39 +02:00
if (is.null(digits) || is.na(digits) || !is.numeric(digits)) {
digits <- 2
}
2020-10-26 12:23:03 +01:00
2020-05-16 13:05:47 +02:00
# round right: percentage(0.4455) and format(as.percentage(0.4455), 1) should return "44.6%", not "44.5%"
x_formatted <- format(round2(as.double(x), digits = digits + 2) * 100,
scientific = FALSE,
2021-05-24 15:29:17 +02:00
digits = max(1, digits),
2020-05-16 13:05:47 +02:00
nsmall = digits,
...)
x_formatted <- paste0(x_formatted, "%")
x_formatted[!grepl(pattern = "^[0-9.,e-]+$", x = x)] <- NA_character_
x_formatted
}
2020-10-26 12:23:03 +01:00
2020-05-16 13:05:47 +02:00
# the actual working part
x <- as.double(x)
if (is.null(digits)) {
# max one digit if undefined
digits <- getdecimalplaces(x, minimum = 0, maximum = 1)
}
format_percentage(structure(.Data = as.double(x),
class = c("percentage", "numeric")),
digits = digits, ...)
}
2020-05-18 13:59:34 +02:00
2021-02-21 20:15:09 +01:00
time_start_tracking <- function() {
2021-05-24 00:06:28 +02:00
pkg_env$time_start <- round(as.double(Sys.time()) * 1000)
2021-02-21 20:15:09 +01:00
}
time_track <- function(name = NULL) {
2021-05-24 00:06:28 +02:00
paste("(until now:", trimws(round(as.double(Sys.time()) * 1000) - pkg_env$time_start), "ms)")
2021-05-18 11:29:31 +02:00
}
2021-04-12 14:24:40 +02:00
# prevent dependency on package 'backports' ----
2021-05-21 20:20:51 +02:00
# these functions were not available in previous versions of R (last checked: R 4.1.0)
2020-05-19 12:08:49 +02:00
# see here for the full list: https://github.com/r-lib/backports
2020-05-19 13:18:01 +02:00
strrep <- function(x, times) {
x <- as.character(x)
2020-10-26 12:23:03 +01:00
if (length(x) == 0L)
2020-05-18 13:59:34 +02:00
return(x)
unlist(.mapply(function(x, times) {
2020-10-26 12:23:03 +01:00
if (is.na(x) || is.na(times))
2020-05-18 13:59:34 +02:00
return(NA_character_)
2020-10-26 12:23:03 +01:00
if (times <= 0L)
2020-05-18 13:59:34 +02:00
return("")
paste0(replicate(times, x), collapse = "")
}, list(x = x, times = times), MoreArgs = list()), use.names = FALSE)
}
2021-05-18 00:53:04 +02:00
trimws <- function(x, which = c("both", "left", "right"), whitespace = "[ \t\r\n]") {
2020-05-19 13:18:01 +02:00
which <- match.arg(which)
mysub <- function(re, x) sub(re, "", x, perl = TRUE)
2021-05-18 00:53:04 +02:00
switch(which,
left = mysub(paste0("^", whitespace, "+"), x),
right = mysub(paste0(whitespace, "+$"), x),
both = mysub(paste0(whitespace, "+$"), mysub(paste0("^", whitespace, "+"), x)))
2020-05-18 13:59:34 +02:00
}
2020-05-19 13:18:01 +02:00
isFALSE <- function(x) {
2020-05-18 13:59:34 +02:00
is.logical(x) && length(x) == 1L && !is.na(x) && !x
}
2020-05-19 13:18:01 +02:00
deparse1 <- function(expr, collapse = " ", width.cutoff = 500L, ...) {
2020-05-19 12:08:49 +02:00
paste(deparse(expr, width.cutoff, ...), collapse = collapse)
}
file.size <- function(...) {
file.info(...)$size
}
file.mtime <- function(...) {
file.info(...)$mtime
}
str2lang <- function(s) {
stopifnot(length(s) == 1L)
2020-09-14 12:21:23 +02:00
ex <- parse(text = s, keep.source = FALSE)
stopifnot(length(ex) == 1L)
ex[[1L]]
}
2020-10-08 11:16:03 +02:00
isNamespaceLoaded <- function(pkg) {
pkg %in% loadedNamespaces()
}
2020-12-22 00:51:17 +01:00
lengths <- function(x, use.names = TRUE) {
2020-12-17 16:22:25 +01:00
vapply(x, length, FUN.VALUE = NA_integer_, USE.NAMES = use.names)
}
2021-04-12 14:24:40 +02:00
2021-05-24 00:06:28 +02:00
if (getRversion() < "3.1") {
2021-04-12 14:24:40 +02:00
# R-3.0 does not contain these functions, set them here to prevent installation failure
# (required for extension of the <mic> class)
2021-04-12 14:24:40 +02:00
cospi <- function(...) 1
sinpi <- function(...) 1
tanpi <- function(...) 1
}
2021-05-21 20:20:51 +02:00
dir.exists <- function (paths) {
x = base::file.info(paths)$isdir
!is.na(x) & x
}