1
0
mirror of https://github.com/msberends/AMR.git synced 2026-02-09 04:32:57 +01:00
Files
AMR/man/as.mic.Rd

150 lines
6.7 KiB
R

% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/mic.R
\docType{data}
\name{as.mic}
\alias{as.mic}
\alias{mic}
\alias{is.mic}
\alias{NA_mic_}
\alias{rescale_mic}
\alias{mic_p50}
\alias{mic_p90}
\alias{droplevels.mic}
\title{Transform Input to Minimum Inhibitory Concentrations (MIC)}
\usage{
as.mic(x, na.rm = FALSE, keep_operators = "all",
round_to_next_log2 = FALSE)
is.mic(x)
NA_mic_
rescale_mic(x, mic_range, keep_operators = "edges", as.mic = TRUE,
round_to_next_log2 = FALSE)
mic_p50(x, na.rm = FALSE, ...)
mic_p90(x, na.rm = FALSE, ...)
\method{droplevels}{mic}(x, as.mic = FALSE, ...)
}
\arguments{
\item{x}{A \link{character} or \link{numeric} vector.}
\item{na.rm}{A \link{logical} indicating whether missing values should be removed.}
\item{keep_operators}{A \link{character} specifying how to handle operators (such as \code{>} and \code{<=}) in the input. Accepts one of three values: \code{"all"} (or \code{TRUE}) to keep all operators, \code{"none"} (or \code{FALSE}) to remove all operators, or \code{"edges"} to keep operators only at both ends of the range.}
\item{round_to_next_log2}{A \link{logical} to round up all values to the next log2 level, that are not either 0.0001, 0.0002, 0.0005, 0.001, 0.002, 0.004, 0.008, 0.016, 0.032, 0.064, 0.125, 0.25, 0.5, 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, or 4096. Values that are already in this list (with or without operators), are left unchanged (including any operators).}
\item{mic_range}{A manual range to rescale the MIC values, e.g., \code{mic_range = c(0.001, 32)}. Use \code{NA} to prevent rescaling on one side, e.g., \code{mic_range = c(NA, 32)}.}
\item{as.mic}{A \link{logical} to indicate whether the \code{mic} class should be kept - the default is \code{TRUE} for \code{\link[=rescale_mic]{rescale_mic()}} and \code{FALSE} for \code{\link[=droplevels]{droplevels()}}. When setting this to \code{FALSE} in \code{\link[=rescale_mic]{rescale_mic()}}, the output will have factor levels that acknowledge \code{mic_range}.}
\item{...}{Arguments passed on to methods.}
}
\value{
Ordered \link{factor} with additional class \code{\link{mic}}, that in mathematical operations acts as a \link{numeric} vector. Bear in mind that the outcome of any mathematical operation on MICs will return a \link{numeric} value.
}
\description{
This transforms vectors to a new class \code{\link{mic}}, which treats the input as decimal numbers, while maintaining operators (such as ">=") and only allowing valid MIC values known to the field of (medical) microbiology.
}
\details{
To interpret MIC values as SIR values, use \code{\link[=as.sir]{as.sir()}} on MIC values. It supports guidelines from EUCAST (2011-2025) and CLSI (2011-2025).
This class for MIC values is a quite a special data type: formally it is an ordered \link{factor} with valid MIC values as \link{factor} levels (to make sure only valid MIC values are retained), but for any mathematical operation it acts as decimal numbers:
\if{html}{\out{<div class="sourceCode">}}\preformatted{x <- random_mic(10)
x
#> Class 'mic'
#> [1] 16 1 8 8 64 >=128 0.0625 32 32 16
is.factor(x)
#> [1] TRUE
x[1] * 2
#> [1] 32
median(x)
#> [1] 26
}\if{html}{\out{</div>}}
This makes it possible to maintain operators that often come with MIC values, such ">=" and "<=", even when filtering using \link{numeric} values in data analysis, e.g.:
\if{html}{\out{<div class="sourceCode">}}\preformatted{x[x > 4]
#> Class 'mic'
#> [1] 16 8 8 64 >=128 32 32 16
df <- data.frame(x, hospital = "A")
subset(df, x > 4) # or with dplyr: df \%>\% filter(x > 4)
#> x hospital
#> 1 16 A
#> 5 64 A
#> 6 >=128 A
#> 8 32 A
#> 9 32 A
#> 10 16 A
}\if{html}{\out{</div>}}
All so-called \link[=groupGeneric]{group generic functions} are implemented for the MIC class (such as \code{!}, \code{!=}, \code{<}, \code{>=}, \code{\link[=exp]{exp()}}, \code{\link[=log2]{log2()}}). Some mathematical functions are also implemented (such as \code{\link[=quantile]{quantile()}}, \code{\link[=median]{median()}}, \code{\link[=fivenum]{fivenum()}}). Since \code{\link[=sd]{sd()}} and \code{\link[=var]{var()}} are non-generic functions, these could not be extended. Use \code{\link[=mad]{mad()}} as an alternative, or use e.g. \code{sd(as.numeric(x))} where \code{x} is your vector of MIC values.
Using \code{\link[=as.double]{as.double()}} or \code{\link[=as.numeric]{as.numeric()}} on MIC values will remove the operators and return a numeric vector. Do \strong{not} use \code{\link[=as.integer]{as.integer()}} on MIC values as by the \R convention on \link{factor}s, it will return the index of the factor levels (which is often useless for regular users).
The function \code{\link[=is.mic]{is.mic()}} detects if the input contains class \code{mic}. If the input is a \link{data.frame} or \link{list}, it iterates over all columns/items and returns a \link{logical} vector.
Use \code{\link[=droplevels]{droplevels()}} to drop unused levels. At default, it will return a plain factor. Use \code{droplevels(..., as.mic = TRUE)} to maintain the \code{mic} class.
With \code{\link[=rescale_mic]{rescale_mic()}}, existing MIC ranges can be limited to a defined range of MIC values. This can be useful to better compare MIC distributions.
For \code{ggplot2}, use one of the \code{\link[=scale_x_mic]{scale_*_mic()}} functions to plot MIC values. They allows custom MIC ranges and to plot intermediate log2 levels for missing MIC values.
\code{NA_mic_} is a missing value of the new \code{mic} class, analogous to e.g. base \R's \code{\link[base:NA]{NA_character_}}.
Use \code{\link[=mic_p50]{mic_p50()}} and \code{\link[=mic_p90]{mic_p90()}} to get the 50th and 90th percentile of MIC values. They return 'normal' \link{numeric} values.
}
\examples{
mic_data <- as.mic(c(">=32", "1.0", "1", "1.00", 8, "<=0.128", "8", "16", "16"))
mic_data
is.mic(mic_data)
# this can also coerce combined MIC/SIR values:
as.mic("<=0.002; S")
# mathematical processing treats MICs as numeric values
fivenum(mic_data)
quantile(mic_data)
all(mic_data < 512)
# rescale MICs using rescale_mic()
rescale_mic(mic_data, mic_range = c(4, 16))
# interpret MIC values
as.sir(
x = as.mic(2),
mo = as.mo("Streptococcus pneumoniae"),
ab = "AMX",
guideline = "EUCAST"
)
as.sir(
x = as.mic(c(0.01, 2, 4, 8)),
mo = as.mo("Streptococcus pneumoniae"),
ab = "AMX",
guideline = "EUCAST"
)
# plot MIC values, see ?plot
plot(mic_data)
plot(mic_data, mo = "E. coli", ab = "cipro")
if (require("ggplot2")) {
autoplot(mic_data, mo = "E. coli", ab = "cipro")
}
if (require("ggplot2")) {
autoplot(mic_data, mo = "E. coli", ab = "cipro", language = "nl") # Dutch
}
}
\seealso{
\code{\link[=as.sir]{as.sir()}}
}
\keyword{datasets}