Plot Symbolic Faces

DESCRIPTION:

USAGE:

faces(x, which = <<see below>>, labels = 1:nrow(x), head = "",  
      max = ncol(x), nrow = <<see below>>, ncol = <<see below>>, 
      fill = T, scale = T, byrow = F) 

REQUIRED ARGUMENTS:

x

matrix of data values with rows representing observations and columns representing variables. Missing values (NA) are allowed.

OPTIONAL ARGUMENTS:

which

the columns of x to be used as the first, second, etc. parameter in the symbolic face. The default is `1:min(15,ncol(x))'. See below for the meaning of the parameters.

labels

vector of character strings to use as labels for the faces (i.e., for the rows of x).

head

character string to use as the heading for the plot.

max

suggested value for the number of rows and columns to go on each page. The default means that there will be ncol(x) rows and ncol(x) columns of faces on each page.

nrow,ncol

specifies exactly the number of rows and columns for the array of plots on each page.

fill

if TRUE, all unused parameters of the face will be set to their nominal (midpoint) value. If FALSE, all features corresponding to unused parameters will not be plotted.

scale

if TRUE, the columns of x will be independently scaled to (0,1). If FALSE no scaling will be done. The data values should then be scaled to the same overall range by some other means; e.g., by scaling the whole of x to the range (0,1).

byrow

logical flag: if TRUE, plots are produced in row-wise order; if FALSE, plots are produced in column-wise order.

SIDE EFFECTS:

DETAILS:

Features involving missing values are not drawn; another strategy would be to assign the mean value of the variable to a missing value.

The feature parameters are: 1-area of face; 2-shape of face; 3-length of nose; 4-location of mouth; 5-curve of smile; 6-width of mouth; 7, 8, 9, 10, 11-location, separation, angle, shape and width of eyes; 12-location of pupil; 13, 14, 15-location, angle and width of eyebrow.

BUGS:

faces resets the graphical parameters mfrow and mfcol to the value c(1,1), even if they had a different value prior to calling faces .

REFERENCES:

Chernoff, H. (1973). The use of Faces to Represent Points in k-Dimensional Space Graphically. Journal of the American Statistical Association, 68, 361-368.

SEE ALSO:

EXAMPLES:

faces(chernoff2[1:9,], 
      head = "Chernoff's Second Example (9 observations)") 
faces(t(cereal.attitude), labels = dimnames(cereal.attitude)[[2]],  
      ncol = 3)