cwhmisc (version 6.0)

plt: Plot depending on switch, Create multiple plots with title and time stamp

Description

- pltCharMat uses output from charMat to plot numerical matrices as characters. - pltRCT executes a (series of) plotting function(s) under the control of some useful switches, may be useful in source. - histRCT creates a (series of) histogram(s), uses pltRCT. - SplomT creates a scatterplot matrix with a) covariances (with script size proportional to size) in the upper triangle, b) histograms (with smoothing) and variable names in the diagonal, and c) scatterplot with smoothes in y and x direction in the lower triangle, stressing high correlations by nearly parallel lines. See figure in other documentation.

Usage

pltCharMat(m,...)
pltRCT(rows, cols, tit="", f = function(x) 0, cex = 1.5,  reset = TRUE, outer = TRUE, oma = c(2, 2, 4, 2), mar = c(4, 4, 2, 1))
histRCT(data, rows = round(sqrt(ncol(data))),  cols = ceiling(ncol(data)/rows), breaks = "Sturges", mainL = deparse(substitute(data)), mainC = colnames(eval(substitute(data))))
SplomT(data, mainL = deparse(substitute(data)), xlabL = "",  hist = "h", adjust = 1, hist.col = trellis.par.get("strip.background")$col[5], cex.diag = 1, h.diag=0.4, colYonX = "red", colXonY = "blue", ...)

Arguments

m

Numerical matrix

tit

Overall title for plot. A vector of one or two elements. If an element is an expression, plotmath will be used.

rows

Number of rows of panels

cols

Number of columns of panels

f

A function to plot the individual plot panels. It can also be a statement sequence {...}.

cex

Font size used for tit

reset

Should previous rows, cols be restored after execution. See note

outer

Passed on to mtext.

oma

Outer margin used in initial par(...).

mar

Lines of margin used in initial par(...).

data

Matrix or dataframe containing data, varibles in columns

breaks

Breaks for histogram

mainL

Label on top of scatterplot matrix or matrix of histograms

mainC

Labels on top of each of the histograms, should be character vector of length = number of columns of data

xlabL

Label for x axis

hist

"h" = histogram, "d" = density curve, "b" = both

adjust

factor to adjust smoothing window for density curve

hist.col

colour for the bars of the histograms

cex.diag

correction factor for font height of correlations and names in the diagonal

h.diag

placement of the variable name in the diagonal panel, =0 means on the lower border, = 0.5 in the middle between lower and upper border

colYonX, colXonY

colour of smoothing lines, y on x and x on y

...

Parameters passed on to upper.panel,lower.panel,diag.panel

Value

These functions are called for their side effect to produce a plot.

WARNING

The sequence of functions contained in f MUST NOT contain any call to postscript, because this would try to open another ps device without closing the old one!

Examples

  x <- rnorm(100); y <- rnorm(100)+1; z <- y+rlnorm(100)
  pltRCT(1,1,f={plot(x,y,xlab="data with trend");
    abline(reg=lm(y~x),lty=2);points(x,z,pch=3)})
  nr <- 100; nc <- 8;
  data <- as.data.frame(matrix(rnorm(nr*nc),nrow=nr,ncol=nc))
  data[,nc]   <- data[,nc-2] + 0.3*data[,nc-1] #generate higher correlations
  data[,nc-1] <- data[,nc-1] + 0.9*data[,nc]
  colnames(data)<-paste("vw",letters[1:nc],sep="")
  SplomT(data,mainL="",hist="d",cex.diag=0.6,hist.col="green")
  SplomT(data,mainL="",hist="b",adjust=0.4,cex.diag = 0.5)
  pltRCT(1, 1, tit="1 by 1 plot", f=plot(y,x-3*y) )
  nr <- 25;  nc <- 16
  pltRCT(1, 2, f={plot(x,y,xlab="my x")
    m <- matrix(rnorm(nr*nc),nrow=25,ncol=nc)
    pltCharMat(m,cex=0.5,col="red")
  })