VFP (version 1.4.3)

precision_plot: Precision Performance Plot of Qualitative Tests.

Description

This function visualizes what is described in the CLSI EP12 guideline for qualitative test with internal continuous response (ICR). The hit rate, i.e. the number of measurements deemed to have a certain condition. The C5 and C95 concentrations will be derived per default by this function but it can be set to any set of hit rates. The histograms representing normal distribution of imprecisions at specific concentrations will be scaled to nicely fit into the plot, i.e. the area under the plot will not be equal to 1.

Usage

precision_plot(
  vfp,
  model.no = NULL,
  cutoff,
  prob = c(0.05, 0.5, 0.95),
  col = c("blue", "black", "red"),
  Cutoff = list(),
  Title = list(),
  Xlabel = list(),
  Ylabel = list(),
  HRLine = list(),
  Legend = FALSE,
  nclass = -1,
  BG = "gray90",
  digits = 3,
  alpha = 0.15,
  alpha2 = 0,
  xlim = NULL,
  col.grid = "white",
  Nrand = 1e+06
)

Arguments

vfp

(VFP) object modeling imprecision over the measuring range

model.no

(integer) specifying the VFP-model to used

cutoff

(numeric) specifying one or two cutoff(s), the latter will implicitly define an equivical zone with implications on how 'prob' will be interpreted (see 'prob' for details)

prob

(numeric) values 0 < x < 1 specifying coverage probability of an respecitive normal distribution at cutoff, in case of two cutoffs all elements of 'prob' < 0.5 will be evaluated in regard to cutoff 1, and all 'prob' > 0.5 in regard to cutoff 2

col

(character) strings specifying colors of the different distributions, which will be plotted semi-transparent using 'alpha1' for specifying the level of transparency (1=opaque, 0=fully transparent)

Cutoff

(list) specifying all parameters of the abline function. Vertical lines representing one or two cutoffs can be specified, the color will be re-used for a label in the upper margin. Set to NULL to omit.

Title

(list) specifying all parameters applicable in function mtext for specifying a main title of the plot

Xlabel

(list) specifying all parameters applicable in function mtext for specifying the X-axis label of the plot

Ylabel

(list) specifying all parameters applicable in function mtext for specifying the Y-axis label of the plot

HRLine

(list) specifying all parameters applicable in lines of the line representing the hit rate developing from 0% to 100%

Legend

(logical) TRUE = a legend is added to the plot

nclass

(integer) number of classes in the histograms representing normal distributions of imprecision at Cx-concentrations, number<10 will lead to automatically determining appropriate numbers per histogram (default)

BG

(character) string specifying a background color

digits

(integer) number of significant digits used to indicated concentrations Cx

alpha

(numeric) value 0<=x<=1 specifying the level of transparency of histograms

alpha2

(numeric) similar to 'alpha' referring to the coverage probability, i.e. setting it to a value < 0 will highlight coverage probabilities in histograms

xlim

(numeric) plotting limits in X-direction

col.grid

(character) string specifying a color name to be used for the grid providing orientation in X- and Y-direction

Nrand

(integer) specifying the number of data points simulated to represent a normal distribution

Author

Andre Schuetzenmeister

Examples

if (FALSE) {
# perform variance component analysis
library(VCA)
data(VCAdata1)
# perform VCA-anaylsis
lst <- anovaVCA(y~(device+lot)/day/run, VCAdata1, by="sample")
# transform list of VCA-objects into required matrix
mat <- get_mat(lst)		# automatically selects "total"
mat
# fit all models batch-wise, the best fitting will be used automatically
res <- fit_vfp(model.no=1:9, Data=mat)
# plot hit and visualize imprecision usign default settings
precision_plot(res, cutoff=20)
# without normal distribution at cutoff do
precision_plot(res, cutoff=20, prob=c(.05, .95), col=c("blue", "red"))
# highlight the proportion > cutoff (hit rate) more 
precision_plot(res, cutoff=20, prob=c(.05, .95), col=c("blue", "red"), alpha2=.5)
# plot with legend
precision_plot(res, cutoff=20, prob=c(.05, .95), col=c("blue", "red"), alpha2=.5, Legend=TRUE)
# use different probabilities and colors
precision_plot(res, cutoff=20, prob=c(.05, .95), col="black", alpha2=.3)

# now using two cutoffs, i.e. with equivocal zone
precision_plot(	res, cutoff=c(17, 19), prob=c(.05, .95), col=c("mediumblue", "red3"), 
					alpha2=.5, HRLine=list(col=c("mediumblue", "red3")))
}