GraphDistatisBoot function

GraphDistatisBoot Plot maps of the factor scores of the observations and their bootstrapped confidence intervals (as confidence ellipsoids or peeled hulls) for a DISTATIS analysis.

GraphDistatisBoot plots maps of the factor scores of the observations from a distatis analysis. GraphDistatisBoot gives a map of the factors scores of the observations plus the boostrapped confidence intervals drawn as "Confidence Ellipsoids" at the percentage level (see parameter percentage).

GraphDistatisBoot(
  FS,
  FBoot,
  axis1 = 1,
  axis2 = 2,
  item.colors = NULL,
  ZeTitle = "Distatis-Bootstrap",
  constraints = NULL,
  nude = FALSE,
  Ctr = NULL,
  lwd = 3.5,
  ellipses = TRUE,
  fill = TRUE,
  fill.alpha = 0.27,
  percentage = 0.95
)

Arguments

• FS: The factor scores of the observations ($res4Splus$F from distatis)
• FBoot: is the bootstrapped factor scores array (FBoot obtained from BootFactorScores or BootFromCompromise)
• axis1: The dimension for the horizontal axis of the plots. (default = 1).
• axis2: The dimension for the vertical axis of the plots (default = 2).
• item.colors: When present, should be a column matrix (dimensions of observations and 1). Gives the color-names to be used to color the plots. Can be obtained as the output of this or the other graph routine. If NULL, prettyGraphs chooses.
• ZeTitle: General title for the plots (default is 'Distatis-Bootstrap').
• constraints: constraints for the axes
• nude: When TRUE do not plot the names of the observations (default is FALSE).
• Ctr: Contributions of each observation. If NULL (default), these are computed from FS.
• lwd: Thickness of the line plotting the ellipse or hull (default = 3.5).
• ellipses: a boolean. When TRUE (default) will plot ellipses (from the car package). When FALSE will plot peeled hulls (from prettyGraphs package).
• fill: when TRUE, fill in the ellipse with color. Relevant for ellipses only.
• fill.alpha: transparency index (a number between 0 and 1) when filling in the ellipses. Relevant for ellipses only (default = .27).
• percentage: A value to determine the percent coverage of the bootstrap partial factor scores to provide ellipse or hull confidence intervals (default = .95).

Returns

• constraints: A set of plot constraints that are returned.

• item.colors: A set of colors for the observations are returned.

Details

The ellipses are plotted using the function dataEllipse() from the package car. The peeled hulls are plotted using the function peeledHulls() from the package prettyGraphs.

Note that, in the current version, the graphs are plotted as R-plots and are not passed back by the function. So the graphs need to be saved "by hand" from the R graphic windows. We plan to improve this in a future version. See also package PTCA4CATA for ggplot2 based graphs.

Examples

# 1. Load the Sort data set from the SortingBeer example
# (available from the DistatisR package)
data(SortingBeer)
# Provide an 8 beers by 10 assessors results of a sorting task
#-----------------------------------------------------------------------------
# 2. Create the set of distance matrices (one distance matrix per assessor)
#    (ues the function DistanceFromSort)
DistanceCube <- DistanceFromSort(Sort)

#-----------------------------------------------------------------------------
# 3. Call the DISTATIS routine with the cube of distance as parameter
testDistatis <- distatis(DistanceCube)
# The factor scores for the beers are in
# testDistatis$res4Splus$F
# the partial factor score for the beers for the assessors are in
#  testDistatis$res4Splus$PartialF
#
# 4. Get the bootstraped factor scores (with default 1000 iterations)
BootF <- BootFactorScores(testDistatis$res4Splus$PartialF)
#-----------------------------------------------------------------------------
# 5. Create the Graphics with GraphDistatisBoot
#
GraphDistatisBoot(testDistatis$res4Splus$F,BootF)

References

The plots are similar to the graphs described in:

Abdi, H., Williams, L.J., Valentin, D., & Bennani-Dosse, M. (2012). STATIS and DISTATIS: Optimum multi-table principal component analysis and three way metric multidimensional scaling. Wiley Interdisciplinary Reviews: Computational Statistics, 4 , 124--167.

Abdi, H., Dunlop, J.P., & Williams, L.J. (2009). How to compute reliability estimates and display confidence and tolerance intervals for pattern classiffers using the Bootstrap and 3-way multidimensional scaling (DISTATIS). NeuroImage, 45 , 89--95.

Abdi, H., & Valentin, D., (2007). Some new and easy ways to describe, compare, and evaluate products and assessors. In D., Valentin, D.Z. Nguyen, L. Pelletier (Eds) New trends in sensory evaluation of food and non-food products. Ho Chi Minh (Vietnam): Vietnam National University-Ho chi Minh City Publishing House. pp. 5--18.

These papers are available from https://personal.utdallas.edu/~herve/

See Also

GraphDistatisAll

GraphDistatisCompromise GraphDistatisPartial

GraphDistatisBoot GraphDistatisRv

distatis

Author(s)

Derek Beaton and Herve Abdi