qtour: Create a tour associated with a mutaframe
In ggobi/cranvas: Interactive statistical graphics based on Qt

Description Usage Arguments Details Value Author(s) Examples

The tourr package is used to create the tour projections. This function creates an R object to manipulate the tour, and all the changes in the tour can be reflected immediately in plots created in cranvas.

1 2	qtour(vars = ~., data, tour_path, aps = 1, fps = 30, rescale = TRUE, sphere = FALSE, ...)

`vars`	variable names to be used in the tour (parsed by `var_names`)
`data`	a mutaframe created by `qdata`
`tour_path`	tour path generator, defaults to 2d grand tour
`aps`	target angular velocity (in radians per second)
`fps`	target frames per second (defaults to 30)
`rescale`	if true, rescale all variables to range [0,1]?
`sphere`	if true, sphere all variables
`...`	ignored

Because the data provided to the tour is a mutaframe, it can listen to changes through listeners on it. Usually these listeners can update the plots which are created from this mutaframe (further, those linked to this mutaframe) on the fly.

Four basic methods can be applied to the object returned by this function (say, tour):

tour$start(): starts the tour (tour projections change successively and are attached to the mutaframe; column names are tour_1, tour_2, ...)
tour$stop(): pauses the tour
tour$slower(): makes the tour slower
tour$faster(): makes it faster.

We can also modify the parameters passed to qtour() on the fly through this object. For example, we can change the type of tour to the guided tour using a syntax like tour$tour_path <- guided_tour(...). All other parameters can be changed similarly, including vars, aps, fps, rescale and sphere.

An object generated by reference classes with signals.

Yihui Xie and Di Cook

library(cranvas)
library(tourr)
data(flea, package = "tourr")

qflea <- qdata(flea, color = species)
flea_tour <- qtour(1:6, data = qflea, tour_path = grand_tour(3))
flea_tour$start()

# Sys.sleep(.5) # need a pause to generate projections
qscatter(proj1, proj2, data = qflea, xlim = c(-0.8, 0.8), ylim = c(-0.8, 
    0.8), xlab = "Projection 1", ylab = "Projection 2")
qscatter(proj1, proj2, edges = matrix(c(1:5, 2:6), ncol = 2, byrow = T), 
    data = qflea, xlim = c(-0.8, 0.8), ylim = c(-0.8, 0.8), xlab = "Projection 1", 
    ylab = "Projection 2")
qparallel(~proj1 + proj2 + proj3, data = qflea)
# qhist(proj1, data = qflea, binwidth = 0.05, xlim = c(-1, 1), ylim =
# c(0, 15))
qhist(proj1, data = qflea, breaks = seq(-1, 1, 0.1), ybreaks = seq(0, 20, 
    5))

flea_tour$pause()

## adjust speed
flea_tour$start()
flea_tour$setSpeed(2)  # speed up (twice as fast)
flea_tour$setSpeed(0.5)  # slow down
flea_tour$faster()
flea_tour$slower()

## change the type of tour
flea_tour$tour_path <- guided_tour(holes, d = 3)
flea_tour$tour_path <- grand_tour(d = 3)
flea_tour$tour_path <- guided_tour(lda_pp(flea$species), d = 3)
flea_tour$tour_path <- grand_tour(d = 3)

## do sphering
flea_tour$sphere <- TRUE

cranvas_off()