reg {gsse401}R Documentation

GSSE 401 - Illustration of regression models

Description

These functions takes arguments defining a regression model with one or two explanatory variables and plots the data, regression model and residuals.

Usage

reg(n.expl, ...)

reg1(true.model, n.points, range.x, regular, x,  ...)

reg2(true.model, n.points, range.x1, range.x2,  
     regular, x1, x2, ...)

Arguments

n.expl the number of explanatory variables. The only values allowed are 1 or 2.
true.model a equation defining the true regression model. Examples of specifications are:
2 + 3*x for one explanatory variable and
3 + 2*x1 - 3*x2 for two explanatory variables
range.x, range.x1, range.x2 a two elements numerical vector with minimum and maximum values for the explanatory variable
n.points number of data points
regular logical. TRUE (the default) indicates that values of the explanatory variable will be regularly spaced between the range of x. FALSE means that values of x will be sampled in the range of x
xvec, xvec1, xvec2 objects or keybord input with values of the explanatory variables. Only used if range.x and range.y are not provided
ask logical. If TRUE the user is prompted before the next plot is shown
... further arguments to be passed to the plot function

Details

The main function reg asks for the number of explanatory variables and then calls reg1 or reg2 for the case of one or two respectively.

The function takes a ``true'' model and values for the explanatory variable(s) and build the response variable from this, adding to the response variable an error with variance equivalent to 20% of its variance.

A linear regression model is fitted and results are displayed in the graphical device.

Value

The function reg1 produces a 2D plot with data against the explanatory variable and adds a line with the fitted regression model.
If the packages scatterplot3d is available, the function reg2 produces a 3D plot with data against the explanatory variables and adds a plane with the regression model fitted to the data.
In both cases a second plot with residuals against fitted value is produced.

Author(s)

Peter J. Diggle p.diggle@lancaster.ac.uk
Paulo Justiniano Ribeiro Jr. p.ribeiro@lancaster.ac.uk.

See Also

lm, scatterplot3d

Examples

reg(1, "1 + 3*x", n.p = 21, range.x = c(0,20), reg = TRUE)
reg(1, "1 + 3*log(x^2)", n.p=25, range.x = c(1, 100), reg=TRUE)
reg(1, "10 - 2*exp(-x/10)", x = 1:30)
##
reg(2, "10 - 5*x1 + 3*x2" , x1 = runif(25), x2 =runif(25))
reg(2, "1+ 3*sqrt(x1) -2*x2^2" , n.p=50, range.x1=c(10, 30),
    range.x2 =c(1, 5))
#
# For an interactive input type:
## Not run: 
reg()
## End(Not run)

[Package gsse401 version 0.0-8 Index]