plotSROC: Draw a Summary ROC curve based on the results of the...

In RichardBirnie/mautils: Utilities to facilitate an automated workflow for meta-analysis

Description

Draw a Summary ROC curve based on the results of the bivariate model

Usage

plotSROC(fit.reitsma, show.data = TRUE, show.prediction = TRUE,
  summary.legend.pos = "bottomright", sroc.legend.pos = "bottomleft", ...)

Arguments

fit.reitsma	An object of class reitsma returned by fitting the Reitsma bivariate model using the function named reitsma in the mada package
show.data	A logical indicating whether or not to show the mean sensitivity and specificity of the individual studies
show.prediction	A logical indicating whether or not draw a prediction region
summary.legend.pos	A character string indicating the position of the legend for the summary point and data points. This is passed to the legend function and must be one of 'bottomleft', 'bottomright' (default), 'topleft', 'topright'.
sroc.legend.pos	A character string indicating the position of the legend for the SROC curve. This is passed to the legend function and must be one of 'bottomleft' (default), 'bottomright', 'topleft', 'topright'.
...	Additional arguments passed to the underlying plot method

Details

This function will draw a summary ROC curve based on a fit of the Reitsma bivariate model. The plot will always include a summary point with a confidence ellipse and a SROC curve. By default the plot will include the data points from the individual studies and a 95 the summary point. These can be excluded by setting the arguments show.data or show.prediction to FALSE

The argument extrapolate may also be set to TRUE or FALSE to indicate whether to extrapolate the SROC curve beyond the observed data points. The default is FALSE indicating no extrapolation

See Also

reitsma-class

RichardBirnie/mautils documentation built on July 12, 2019, 8:56 p.m.