Learn R Programming

altmeta (version 4.1)

plot.meta.dt: Plot for Meta-Analysis of Diagnostic Tests

Description

Visualizes meta-analysis of diagnostic tests by presenting summary results, such as ROC (receiver operating characteristic) curve, overall sensitivity and overall specificity (1 \(-\) specificity), and their confidence and prediction regions.

Usage

# S3 method for meta.dt
plot(x, add = FALSE, xlab, ylab, alpha,
     studies = TRUE, cex.studies, col.studies, pch.studies,
     roc, col.roc, lty.roc, lwd.roc, weight = FALSE,
     eqline, col.eqline, lty.eqline, lwd.eqline,
     overall = TRUE, cex.overall, col.overall, pch.overall,
     confid = TRUE, col.confid, lty.confid, lwd.confid,
     predict = FALSE, col.predict, lty.predict, lwd.predict, ...)

Value

None.

Arguments

x

an object of class "meta.dt" created by the function meta.dt.

add

a logical value indicating if the plot is added to an already existing plot.

xlab

a label for the x axis; the default is "1 - Specificity".

ylab

a label for the y axis; the default is "Sensitivity".

alpha

a numeric value specifying the statistical significance level for the confidence and prediction regions. If not specified, the plot uses the significance level stored in x (i.e., x$alpha).

studies

a logical value indicating if the individual studies are presented in the plot.

cex.studies

the size of points representing individual studies (the default is 1).

col.studies

the color of points representing individual studies (the default is "black").

pch.studies

the symbol of points representing individual studies (the default is 1, i.e., circle).

roc

a logical value indicating if the ROC curve is presented in the plot. The default is TRUE for the summary ROC approach (x$method = "s.roc") and is FALSE for the bivariate (generalized) linear mixed model (x$method = "biv.lmm" or "biv.glmm").

col.roc

the color of the ROC curve (the default is "black").

lty.roc

the line type of the ROC curve (the default is 1, i.e., solid line).

lwd.roc

the line width of the ROC curve (the default is 1).

weight

a logical value indicating if the weighted (TRUE) or unweighted (FALSE, the default) regression is used for the summary ROC approach (when x$method is "s.roc").

eqline

a logical value indicating if the line of sensitivity equaling to specificity is presented in the plot.

col.eqline

the color of the equality line (the default is "black").

lty.eqline

the type of the equality line (the default is 4, i.e., dot-dash line).

lwd.eqline

the width of the equality line (the default is 1).

overall

a logical value indicating if the overall sensitivity and overall specificity are presented in the plot. This and the following arguments are used for the bivariate (generalized) linear mixed model (x$method = "biv.lmm" or "biv.glmm").

cex.overall

the size of the point representing the overall sensitivity and overall specificity (the default is 1).

col.overall

the color of the point representing the overall sensitivity and overall specificity (the default is "black").

pch.overall

the symbol of the point representing the overall sensitivity and overall specificity (the default is 15, i.e., filled square).

confid

a logical value indicating if the confidence region of the overall sensitivity and overall specificity is presented in the plot.

col.confid

the line color of the confidence region (the default is "black").

lty.confid

the line type of the confidence region (the default is 2, i.e., dashed line).

lwd.confid

the line width of the confidence region (the default is 1).

predict

a logical value indicating if the prediction region of the overall sensitivity and overall specificity is presented in the plot.

col.predict

the line color of the prediction region (the default is "black").

lty.predict

the line type of the prediction region (the default is 3, i.e., dotted line).

lwd.predict

the line width of the prediction region (the default is 1).

...

other arguments that can be passed to the function plot.default.

See Also

meta.dt, print.meta.dt