panel.qqmath.fitted: Panel function for 'qqmathFitted'
In christopherggreen/cggmisc: Christopher G Green's Miscellaneous Functions

Description Usage Arguments Details Author(s) See Also Examples

This is the default panel function for qqmathFitted.

    panel.qqmath.fitted(x, 
        f.value = NULL, 
        distribution = qnorm, 
        quantile.type = 7, 
        groups = NULL, 
        subscripts = NULL, 
        fitted.args = NULL, 
        qqstyle = NULL, 
        envelope = TRUE, 
        rdist = rnorm, 
        rdist.args = NULL, 
        sig.level = 0.95, 
        mc.samples = 100, 
        seed = .Random.seed, 
        envstyle = NULL, 
        add.qqline = c("none", "lm", "lmRob"), 
        add.qqline.par, 
        id.n = 3, 
        labelargs = FALSE, 
        verbose = FALSE, ...)

`x`	vector of numeric data to be plotted in this panel
`f.value`	Vector of probabilities at which to compute quantiles. By default, `ppoints` is used to pick sensible values, so you generally don't need to specify `f.value`. See `qqmath` and `prepanel.qqmath.fitted` for more details.
`distribution`	function to compute quantiles from the reference distribution. Can be a named function (e.g., `qnorm`), a character string containing the name of a function (e.g., `"qnorm"`), or an anonymous function.
`quantile.type`	Numeric value, passed to `quantile`. Controls the type of sample quantiles that will be taken from the observations `x`. See the help page for `quantile` for more detail.
`groups`	Standard `lattice` argument defining data groups. See the help page for `xyplot`.
`subscripts`	Standard `lattice` argument for use with `groups`. See the help page for `xyplot`.
`fitted.args`	This is a list of fitted parameters for the reference distribution. The user should not specify these directly. Use the `distribution.fit` argument to `qqmathFitted` to provide a function for estimating distributional parameters.
`qqstyle`	List of graphical parameters to use for plotting the observed and reference quantiles.
`envelope`	If TRUE, plot a confidence region for the Q-Q plot via simulation.
`rdist`	The distribution from which to draw random numbers for the confidence region simulation. It should correspond to the distribution assumed in `distribution` and `distribution.fit`. Just like `distribution`, `rdist` can be specified as a named function, a character string containing the name of a function, or an anonymous function.
`rdist.args`	A list containing additional arguments to the function specified in `rdist` (other than the parameters return by `distribution.fit`).
`sig.level`	Significance level to use for the confidence region. Default is a 95% condifence interval.
`mc.samples`	Number of random samples to use to construct the confidence region.
`seed`	Seed to use for the random number generator. If specified, `.Random.seed` will be set to `seed` prior to generation of the simulation envelopes. This is useful to do for reproducibility of results.
`envstyle`	List of graphical parameters to use for plotting the confidence region boundaries.
`add.qqline`	Controls whether a fitted line should be added to the plot. The default option, `"none"`, does not add a line. The `"lm"` option adds a linear regression line (regressing observations on quantiles from the reference distribution) using `"lm"`. The `"lmRob"` option adds a robust regression line, and requires the `robust` library.
`add.qqline.par`	Optional list of parameters controlling the appearance of the fitted line.
`id.n`	Number of outliers to label in each panel. Default is 3. Use `id.n = 0` or `id.n = FALSE` to omit labels for outliers. The outliers that will be flagged are the `id.n` largest in absolute value. The labels correspond to the index of each point in the provided vector of observations.
`labelargs`	Controls whether the fitted distributional parameters are added to each panel. Can be a logical value or a character value. If FALSE, no labels are added. If TRUE, labels are taken from the names of `rdist.args`. If a character vector is provided, those strings are used to subset the names of `rdist.args`. The panel function `panel.qqmath.fitted` will prepend each label to the corresponding distributional parameter estimate, rounded to two decimal places. `labelargs` is not supported when groups are used due to the difficulty of automatically producing readable labels for all groups in each panel. The format of the labels is controlled by the lattice setting `add.text`, and can be influenced by changing that setting. Additional settings can be changed by passing arguments to `gpar` from the `grid` package via the `...` mechanism.
`verbose`	If TRUE, additional debugging information will be printed.
`...`	Further arguments, often graphical parameters, passed to internal plotting functions and eventually on to `panel.superpose`, `panel.xyplot`, and `gpar`.

qqmathFitted is a modified version of qqmath from Deepayan Sarkar's lattice package. It adds the ability to fit the reference distribution to the data during the plotting procedure, as well as the ability to simulation envelopes. The latter ability is based on code appearing in the function lmfmResQQPlot from the robust package maintained by Kjell Konis.

prepanel.qqmath.fitted does the actual distributional fitting for qqmathFitted. The results of the fitting are stored in the frame used by qqmathFitted so that they do not need to be recomputed by panel.qqmath.fitted.

Christopher G. Green christopher.g.green@gmail.com for panel.qqmath.fitted. panel.qqmath.fitted is a modified version of panel.default.qqmath from the lattice package, written by Deepayan Sarkar. The plotting code for simulation envelopes is based on code appearing in the lmfmResQQPlot function from the robust package written by Kjell Konis et al.

qqmathFitted, prepanel.qqmath.fitted, qqmath, panel.qqmath