R: Create an empty log-probability plot

lpplot {heR.Misc}

R Documentation

Create an empty log-probability plot

Description

This function creates an empty log-probability plot with nice tics and, optionally, gridlines

Usage

lpplot(xlim, ylim, grid=TRUE, grid.lty="solid", grid.col="lightgray",
       main = "Log-Probability Plot",
       xlab = "Standard Normal Cumulative Probability (
       ylab = "Sample Quantiles",
       bty="o", axes = TRUE, cex.main=1, cex.lab=1, cex.axis=1, las=1,
       tck.minor=-0.009, tck.major=-0.017,
       xtic.minor=FALSE, ...)

Arguments

`ylim`	a two-element vector giving the y-axis limits, defaulting to c(1,1000)
`grid`	logical, whether to draw grid lines or not, defaulting to `TRUE`
`grid.lty`	line type to use for grid lines, defaulting to "dotted"
`main,xlab,ylab`	The plot title and axis labels, defaulting to appropriate values
`cex.main,cex.lab,cex.axis`	magnification factors for the main title, axis label, and axis tic labels, respectively
`las`	specification for the orientation (style) of axis tic labels, can have values of 1:4, defaults to horizontal labels for both axes, `las=1`
`tck.minor, tck.major`	specifications for the length (and direction) of major and minor axis tic marks in units of a fraction of the plotting region; negative values make the tics point away from the plot (i.e., the typical direction)
`xtic.minor`	logical, whether the tic marks for the x-axis will be the length of minor tics or not, defaults to `FALSE`, meaning that they will be the length of major tics.
`...`	additional graphics parameters for the `axis` function

Details

The log-probability plot is a type of Q-Q (quantile-quantile) plot commonly used to judge how data conform to a specific model or data vector. The y-axis is a log-scale, so that sample points appear as if they were log-transformed. They are plotted against corresponding normal quantiles on the x-axis.

If the curve appears as a straight line on this plot, then the data are lognormally distributed, or, in other words, that the log of the data are normally distributed.

The x-axis, corresponding to normal quantiles, is labeled with cumulative normal probabilities, and the y-axis is labeled with untransformed data (sample) quantiles.

Use the functions add.lp.data, add.lp.bin, and add.lp.lnorm to add points or lines to a log-probablity plot produced by this function.

Use the qqnorm function for a normal quantile-quantile plot.

See the axis function for explanations of graphical parameters associated with axes.

Value

Invoked for the side effect of creating an empty log-probability plot

Author(s)

Neil E. Klepeis