1
0
mirror of https://github.com/msberends/AMR.git synced 2025-07-08 20:41:58 +02:00

(v1.5.0.9022) mo properties speed improvement

This commit is contained in:
2021-02-21 22:56:35 +01:00
parent 5ef8cb41a7
commit 062c49fca1
19 changed files with 76 additions and 47 deletions

View File

@ -112,7 +112,7 @@ check_dataset_integrity <- function() {
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 names was loaded _after_ the AMR package, preventing the AMR package from working correctly. Please load the AMR package last.")
"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.")
invisible(TRUE)
}
@ -538,7 +538,7 @@ meet_criteria <- function(object,
if (!is.null(pkg_env$meet_criteria_error_txt)) {
error_txt <- pkg_env$meet_criteria_error_txt
pkg_env$meet_criteria_error_txt <- NULL
stop(error_txt, call. = FALSE)
stop(error_txt, call. = FALSE) # don't use stop_() here, pkg may not be loaded yet
}
pkg_env$meet_criteria_error_txt <- NULL
@ -867,7 +867,13 @@ font_grey <- function(..., collapse = " ") {
try_colour(..., before = "\033[38;5;249m", after = "\033[39m", collapse = collapse)
}
font_grey_bg <- function(..., collapse = " ") {
try_colour(..., before = "\033[48;5;254m", after = "\033[49m", collapse = collapse)
if (tryCatch(rstudioapi::getThemeInfo()$dark == TRUE, error = function(e) FALSE)) {
# similar to HTML #444444
try_colour(..., before = "\033[48;5;238m", after = "\033[49m", collapse = collapse)
} else {
# similar to HTML #eeeeee
try_colour(..., before = "\033[48;5;254m", after = "\033[49m", collapse = collapse)
}
}
font_green_bg <- function(..., collapse = " ") {
try_colour(..., before = "\033[42m", after = "\033[49m", collapse = collapse)

25
R/mo.R
View File

@ -178,13 +178,6 @@ as.mo <- function(x,
return(set_clean_class(x, new_class = c("mo", "character")))
}
if (tryCatch(all(x == "" | gsub(".*(unknown ).*", "unknown name", tolower(x), perl = TRUE) %in% MO_lookup$fullname_lower, na.rm = TRUE)
& isFALSE(Becker)
& isFALSE(Lancefield), error = function(e) FALSE)) {
# to improve speed, special case for taxonomically correct full names (case-insensitive)
return(MO_lookup[match(gsub(".*(unknown ).*", "unknown name", tolower(x), perl = TRUE), MO_lookup$fullname_lower), "mo", drop = TRUE])
}
# start off with replaced language-specific non-ASCII characters with ASCII characters
x <- parse_and_convert(x)
# replace mo codes used in older package versions
@ -198,6 +191,13 @@ as.mo <- function(x,
x[trimws2(x) %like% translate_AMR("no .*growth", language = language)] <- NA_character_
x[trimws2(x) %like% paste0("^(", translate_AMR("no|not", language = language), ") [a-z]+")] <- "UNKNOWN"
uncertainty_level <- translate_allow_uncertain(allow_uncertain)
if (tryCatch(all(x == "" | gsub(".*(unknown ).*", "unknown name", tolower(x), perl = TRUE) %in% MO_lookup$fullname_lower, na.rm = TRUE)
& isFALSE(Becker)
& isFALSE(Lancefield), error = function(e) FALSE)) {
# to improve speed, special case for taxonomically correct full names (case-insensitive)
return(MO_lookup[match(gsub(".*(unknown ).*", "unknown name", tolower(x), perl = TRUE), MO_lookup$fullname_lower), "mo", drop = TRUE])
}
if (!is.null(reference_df)
&& check_validity_mo_source(reference_df)
@ -481,11 +481,11 @@ exec_as.mo <- function(x,
# Fill in fullnames and MO codes at once
known_names <- tolower(x_backup) %in% MO_lookup$fullname_lower
x[known_names] <- MO_lookup[match(tolower(x_backup)[known_names], MO_lookup$fullname), property, drop = TRUE]
x[known_names] <- MO_lookup[match(tolower(x_backup)[known_names], MO_lookup$fullname_lower), property, drop = TRUE]
known_codes <- toupper(x_backup) %in% MO_lookup$mo
x[known_codes] <- MO_lookup[match(toupper(x_backup)[known_codes], MO_lookup$mo), property, drop = TRUE]
already_known <- known_names | known_codes
# now only continue where the right taxonomic output is not already known
if (any(!already_known)) {
x_known <- x[already_known]
@ -984,7 +984,6 @@ exec_as.mo <- function(x,
g.x_backup_without_spp %pm>% substr(1, x_length / 2),
".* ",
g.x_backup_without_spp %pm>% substr((x_length / 2) + 1, x_length))
print(x_split)
found <- lookup(fullname_lower %like_case% x_split,
haystack = data_to_check)
if (!is.na(found)) {
@ -1977,13 +1976,13 @@ parse_and_convert <- function(x) {
if (NCOL(x) > 2) {
stop("a maximum of two columns is allowed", call. = FALSE)
} else if (NCOL(x) == 2) {
# support Tidyverse selection like: df %pm>% select(colA, colB)
# support Tidyverse selection like: df %>% select(colA, colB)
# paste these columns together
x <- as.data.frame(x, stringsAsFactors = FALSE)
colnames(x) <- c("A", "B")
x <- paste(x$A, x$B)
} else {
# support Tidyverse selection like: df %pm>% select(colA)
# support Tidyverse selection like: df %>% select(colA)
x <- as.data.frame(x, stringsAsFactors = FALSE)[[1]]
}
}
@ -1991,6 +1990,8 @@ parse_and_convert <- function(x) {
parsed <- iconv(x, to = "UTF-8")
parsed[is.na(parsed) & !is.na(x)] <- iconv(x[is.na(parsed) & !is.na(x)], from = "Latin1", to = "ASCII//TRANSLIT")
parsed <- gsub('"', "", parsed, fixed = TRUE)
parsed <- gsub(" +", " ", parsed, perl = TRUE)
parsed <- trimws(parsed)
}, error = function(e) stop(e$message, call. = FALSE)) # this will also be thrown when running `as.mo(no_existing_object)`
parsed
}

View File

@ -694,22 +694,22 @@ mo_property <- function(x, property = "fullname", language = get_locale(), ...)
mo_validate <- function(x, property, language, ...) {
check_dataset_integrity()
if (tryCatch(all(x[!is.na(x)] %in% MO_lookup$mo) & length(list(...)) == 0, error = function(e) FALSE)) {
# special case for mo_* functions where class is already <mo>
return(MO_lookup[match(x, MO_lookup$mo), property, drop = TRUE])
}
dots <- list(...)
Becker <- dots$Becker
if (is.null(Becker)) {
if (is.null(Becker) | property %in% c("kingdom", "phylum", "class", "order", "family", "genus")) {
Becker <- FALSE
}
Lancefield <- dots$Lancefield
if (is.null(Lancefield)) {
if (is.null(Lancefield) | property %in% c("kingdom", "phylum", "class", "order", "family", "genus")) {
Lancefield <- FALSE
}
has_Becker_or_Lancefield <- Becker %in% c(TRUE, "all") | Lancefield %in% c(TRUE, "all")
if (tryCatch(all(x[!is.na(x)] %in% MO_lookup$mo) & !has_Becker_or_Lancefield, error = function(e) FALSE)) {
# special case for mo_* functions where class is already <mo>
return(MO_lookup[match(x, MO_lookup$mo), property, drop = TRUE])
}
# try to catch an error when inputting an invalid argument
# so the 'call.' can be set to FALSE
tryCatch(x[1L] %in% MO_lookup[1, property, drop = TRUE],
@ -722,8 +722,13 @@ mo_validate <- function(x, property, language, ...) {
# because it's already a valid MO
x <- exec_as.mo(x, property = property, initial_search = FALSE, language = language, ...)
} else if (!all(x %in% MO_lookup[, property, drop = TRUE])
| Becker %in% c(TRUE, "all")
| Lancefield %in% c(TRUE, "all")) {
| has_Becker_or_Lancefield) {
accepted_args <- names(as.list(args("as.mo")))
accepted_args <- accepted_args[!accepted_args %in% c("", "...", "x", "property")]
stop_if(!all(names(dots) %in% names(as.list(args("as.mo")))),
"invalid argument(s): ", vector_and(names(dots)[!names(dots) %in% names(as.list(args("as.mo")))], quotes = "'"),
".\nAccepted arguments are ", vector_and(accepted_args, quotes = "'"), ".",
call = FALSE)
x <- exec_as.mo(x, property = property, language = language, ...)
}

Binary file not shown.