AMR/R/misc.R

95 lines
3.7 KiB
R
Executable File

# ==================================================================== #
# TITLE #
# Antimicrobial Resistance (AMR) Analysis #
# #
# AUTHORS #
# Berends MS (m.s.berends@umcg.nl), Luz CF (c.f.luz@umcg.nl) #
# #
# LICENCE #
# This program is free software; you can redistribute it and/or modify #
# it under the terms of the GNU General Public License version 2.0, #
# as published by the Free Software Foundation. #
# #
# This program is distributed in the hope that it will be useful, #
# but WITHOUT ANY WARRANTY; without even the implied warranty of #
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the #
# GNU General Public License for more details. #
# ==================================================================== #
# No export, no Rd
addin_insert_in <- function() {
rstudioapi::insertText(" %in% ")
}
# No export, no Rd
addin_insert_like <- function() {
rstudioapi::insertText(" %like% ")
}
# No export, no Rd
percent <- function(x, round = 1, force_zero = FALSE, ...) {
val <- base::round(x * 100, digits = round)
if (force_zero == TRUE & any(val == as.integer(val) & !is.na(val))) {
val[val == as.integer(val)] <- paste0(val[val == as.integer(val)], ".", strrep(0, round))
}
pct <- base::paste0(val, "%")
pct[pct == "NA%"] <- NA_character_
pct
}
check_available_columns <- function(tbl, col.list, info = TRUE) {
# check columns
col.list <- col.list[!is.na(col.list)]
names(col.list) <- col.list
col.list.bak <- col.list
# are they available as upper case or lower case then?
for (i in 1:length(col.list)) {
if (toupper(col.list[i]) %in% colnames(tbl)) {
col.list[i] <- toupper(col.list[i])
} else if (tolower(col.list[i]) %in% colnames(tbl)) {
col.list[i] <- tolower(col.list[i])
} else if (!col.list[i] %in% colnames(tbl)) {
col.list[i] <- NA
}
}
if (!all(col.list %in% colnames(tbl))) {
if (info == TRUE) {
warning('These columns do not exist and will be ignored: ',
col.list.bak[!(col.list %in% colnames(tbl))] %>% toString(),
'.\nTHIS MAY STRONGLY INFLUENCE THE OUTCOME.',
immediate. = TRUE,
call. = FALSE)
}
}
col.list
}
# Coefficient of variation (CV)
cv <- function(x, na.rm = TRUE) {
stats::sd(x, na.rm = na.rm) / base::abs(base::mean(x, na.rm = na.rm))
}
# Coefficient of dispersion, or coefficient of quartile variation (CQV).
# (Bonett et al., 2006: Confidence interval for a coefficient of quartile variation).
cqv <- function(x, na.rm = TRUE) {
fives <- stats::fivenum(x, na.rm = na.rm)
(fives[4] - fives[2]) / (fives[4] + fives[2])
}
# show bytes as kB/MB/GB
# size_humanreadable(123456) # 121 kB
# size_humanreadable(12345678) # 11.8 MB
size_humanreadable <- function(bytes, decimals = 1) {
bytes <- bytes %>% as.double()
# Adapted from:
# http://jeffreysambells.com/2012/10/25/human-readable-filesize-php
size <- c('B','kB','MB','GB','TB','PB','EB','ZB','YB')
factor <- floor((nchar(bytes) - 1) / 3)
# added slight improvement; no decimals for B and kB:
decimals <- rep(decimals, length(bytes))
decimals[size[factor + 1] %in% c('B', 'kB')] <- 0
out <- paste(sprintf(paste0("%.", decimals, "f"), bytes / (1024 ^ factor)), size[factor + 1])
out
}