plot.gvf.fit: Diagnostic Plots for Fitted GVF Models
In DiegoZardetto/ReGenesees: R Evolved Generalized Software for Sampling Estimates and Errors in Surveys

plot.gvf.fit

R Documentation

Diagnostic Plots for Fitted GVF Models

Description

This function provides basic diagnostic plots for fitted GVF model(s).

Usage

## S3 method for class 'gvf.fit'
plot(x, which.more = 1:3, id.n = 3, labels.id = names(residuals(x)),
     cex.id = 0.75, label.pos = c(4, 2), cex.caption = 1, Main = NULL, ...)

## S3 method for class 'gvf.fits'
plot(x, which.more = NULL, id.n = 3, labels.id = names(residuals(x)),
     cex.id = 0.75, label.pos = c(4, 2), cex.caption = 1, Main = NULL, ...)

## S3 method for class 'gvf.fit.gr'
plot(x, which.more = 1:3, id.n = 3, labels.id = NULL,
     cex.id = 0.75, label.pos = c(4, 2), cex.caption = 1, ...)

## S3 method for class 'gvf.fits.gr'
plot(x, which.more = NULL, id.n = 3, labels.id = NULL,
     cex.id = 0.75, label.pos = c(4, 2), cex.caption = 1, ...)

Arguments

`x`	An object containing one or more fitted GVF models (see ‘Usage’ for the allowed classes).
`which.more`	Select additional plots beyond the default one (‘Observed vs Fitted’). Can be any subset of vector `1:6` with up to three elements.
`id.n`	Number of points to be initially labelled in each plot, starting with the most extreme.
`labels.id`	Vector of labels, from which the labels for extreme points will be chosen. `NULL` uses observation numbers.
`cex.id`	Magnification of point labels.
`label.pos`	Positioning of labels, for the left half and right half of the graph(s) respectively.
`cex.caption`	Controls the size of `caption`.
`Main`	Optional string to be added to automatic plot titles.
`...`	Other parameters to be passed through to plotting functions.

Details

Diagnostic plots can be useful both for assessing the goodness of a GVF model fit qualitatively, and for selecting the “best” GVF model among different alternatives.

This function can provide any of the following 7 plots:

 (0)  'Observed vs Fitted'
 (1)  'Residuals vs Fitted'
 (2)  'Normal Q-Q'
 (3)  'Scale-Location'
 (4)  'Cook's distance'
 (5)  'Residuals vs Leverage'
 (6)  'Cook's distances vs Leverage/(1-Leverage)'

The ‘Residuals vs Fitted’ plot is special in that it will be always provided: this explains its zero-th order in the list above. The rest of the list, namely plots 1:6, exactly matches the numbering convention of function plot.lm.

Additional plots - beyond ‘Residuals vs Fitted’ - can be requested through argument which.more. Any subset of 1:6 is allowed, provided its length does not exceed 3. Therefore, at most 4 plots will be generated simultaneously.

Note that the default behaviour of this function do depend on whether input object x stores one or more than one fitted GVF models. In the first case, plots 0:3 will be returned in a multiple plot with a 2x2 layout. In the second case, only the default plot number 0 will be returned, opening a new graphics frame for each different GVF model.

Argument id.n specifies how many points have to be labelled, starting with the most extreme in terms of residuals: this applies to all plots.

Argument Main is expected to be seldom (if ever) useful: its main purpose is programming consistency at a deeper level.

All the other arguments have the same meaning as in function plot.lm.

Author(s)

Diego Zardetto

References

See plot.lm and references therein.

Examples

# Load example data:
data(AF.gvf)

# Inspect available estimates and errors of counts:
str(ee.AF)

# List available registered GVF models:
GVF.db


############################################
# Diagnostic plots for fitted GVF model(s) #
############################################

## (A) Plots of a *single* fitted GVF model ##
# Fit example data to registered GVF model number one:
m <- fit.gvf(ee.AF, 1)

## Default call yields 4 plots:
plot(m)

# Play with argument 'which.more':
## which.more = NULL yields the "Observed vs Fitted" plot only:
plot(m, which.more = NULL)

## which.more = 1 adds the "Residuals vs Fitted" plot:
plot(m, which.more = 1)

## subsets of 1:6 with length <= 3 are allowed:
plot(m, which.more = c(1:2,4))

# Just for illustration, play with other parameters:
plot(m, id.n = 6, col = "blue", pch = 20)


## (B) Plots of *many* fitted GVF models ##
# Fit example data to registered GVF models for frequencies (i.e. number 1:3):
mm <- fit.gvf(ee.AF, 1:3)

## Default call yields the "Observed vs Fitted" plot reported separately for
## each model in subsequent graphics frames:
plot(mm)

# Play with argument 'which.more':
## which.more = 1:3 yields subsequent 2x2 plots:
plot(mm, which.more = 1:3)

## again, subsets of 1:6 are allowed:
plot(mm, which.more = 1)


######################################################
# Diagnostic plots for "grouped" fitted GVF model(s) #
######################################################
# We have at our disposal the following survey design object on household data:
exdes

# Use function svystat to prepare "grouped" estimates and errors of counts
# to be fitted separately (here groups are regions):
ee <- svystat(exdes, y=~ind, by=~age5c:marstat:sex, combo=3, group=~regcod)
lapply(ee, head)

## (C) Plots of a *single* GVF model fitted to different groups ##
# Fit registered GVF model number one separately inside groups:
m.g <- fit.gvf(ee, 1)

## Default call yields 4 plots reported separately for each group
## in subsequent graphics frames:
plot(m.g)

## Play with argument 'which.more' to select different plots:
plot(m.g, which.more = c(1:2,4))


## (D) Plots of *many* GVF model fitted to different groups ##
# Fit all registered GVF models for frequencies separately inside groups:
mm.g <- fit.gvf(ee, 1:3)

## Default call yields the "Residuals vs Fitted" plot reported separately
## for each group in subsequent graphics frames:
plot(mm.g)

## Play with argument 'which.more' to add more plots:
plot(mm.g, which.more = 1:3, id.n = 6, col = "blue", pch = 20)

DiegoZardetto/ReGenesees documentation built on Dec. 16, 2024, 2:03 p.m.