fitPath {condvis2}

Constructs tours of data space based on fits.

Description

Constructs tours of data space based on fits.

Usage

lofPath(
  data,
  fits,
  length = 10,
  reorder = TRUE,
  conditionvars = NULL,
  predictArgs = NULL,
  response = NULL,
  ...
)

diffitsPath(
  data,
  fits,
  length = 10,
  reorder = TRUE,
  conditionvars = NULL,
  predictArgs = NULL,
  ...
)

hiresponsePath(
  data,
  response = NULL,
  length = 10,
  reorder = TRUE,
  conditionvars = NULL,
  ...
)

loresponsePath(
  data,
  response = NULL,
  length = 10,
  reorder = TRUE,
  conditionvars = NULL,
  ...
)

Arguments

data

A dataframe

fits

A model fit or list of fits

length

Path length, defaults to 10

reorder

If TRUE (default) uses DendSer to reorder the path dser

conditionvars

A vector of variable names. The returned tour is for this subset of variables.

predictArgs

Extra inputs to CVpredict

response

The name of the response variable

...

ignored

Value

A dataframe with the path

Functions

• lofPath(): Constructs a tour of data space showing biggest absolute residuals from fits.

• diffitsPath(): Constructs a tour of data space showing biggest differences in fits.

• hiresponsePath(): Constructs a tour of data space showing high (numeric) response values

• loresponsePath(): Constructs a tour of data space showing low (numeric) response values

Examples

fit1 <- lm(mpg ~ wt+hp+am, data=mtcars)
fit2 <- lm(mpg ~ wt, data=mtcars)
lofPath(mtcars,fit1, response="mpg")
diffitsPath(mtcars,list(fit1,fit2))