plot.logitreg {analogue} R Documentation

## Produces plots of analogue logistic regression models

### Description

Draws the fitted logistic regression function describing the posterior probability that two sites are analogues conditional upon the dissimilarity between the two samples. Confidence intervals are also computed and displayed if requested.

### Usage

```## S3 method for class 'logitreg'
plot(x, group = "all", npred = 100,
conf.int = 0.9, conf.type = c("none", "polygon", "lines"),
xlab = expression(D[ij]), ylab = "Pr (A+ | d)",
rug = TRUE, ticksize = 0.02,
col = "red", ref.col = "lightgrey",
lwd = 2, conf.lwd = 1, conf.lty = "dashed",
```

### Arguments

 `x` object to plot; an object of class `"logitreg"`, usually the result of a call to `logitreg`. `group` The group to plot the logit model for. Can be one of the group labels or `"Combined"` to draw the individual logit models. Alternatively, and the default, is to use `"all"`, which divides the plotting region into the required number of plotting regions and draws all the fitted curves. `npred` number of points at which the fitted curves are evaluated for plotting purposes. `conf.int` numeric; the confidence interval required. `conf.type` character; how should the confidence interval be drawn. Default is not to draw the confidence interval. `xlab, ylab` character; the x and y axis labels. `rug` logical; should rug plots be drawn? `ticksize` The size of the tick marks used in rug plots. `col` The colour in which to draw the line representing the fitted values. `ref.col` The colour of the reference lines drawn at 0 and 1. `lwd` The line width in which to draw the line representing the fitted values. `conf.lwd, conf.lty` Line width and line type for the confidence interval. Only used if `conf.type = "lines"` is specified. `shade` The colour for the fill and border of the confidence interval if `conf.type = "polygon"` is specified. `...` arguments passed on to `plot`.

### Value

A plot on the current device.

### Author(s)

Gavin L. Simpson

`logitreg` for an example, `roc`