R image function - error with color/value scaling

Question

I have a question about the R function image in the biwavelet package. This function is used to plot correlation values as the z in the image function. I'm plotting perfectly correlated data sets (R-squared=1 as my z-value). When the image is plotted there is some variation in the colored boxes. I'm confused by this because all the values in my z matrix are 1. Any ideas why this could be happening, and how to fix it?

image(x$t, yvals, t(zvals), zlim = zlim, xlim = xlim, ylim = rev(ylim), 
      xlab = "Time", ylab = "Period", yaxt = "n", xaxt = "n", col = fill.colors)

zvals is a matrix of all 1s

UPDATE: based on request below

 x$t
 [1] 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986     1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
 [29] 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

yvals
[1] 1.046901 1.130235 1.213568 1.296901 1.380235 1.463568 1.546901 1.630235   1.713568 1.796901 1.880235 1.963568 2.046901 2.130235 2.213568
[16] 2.296901 2.380235 2.463568 2.546901 2.630235 2.713568 2.796901 2.880235 2.963568 3.046901 3.130235 3.213568 3.296901 3.380235 3.463568
[31] 3.546901 3.630235 3.713568 3.796901

str(zvals)
num [1:34, 1:40] 1 1 1 1 1 1 1 1 1 1 ...

Above are the actual data. Below is an example, but plots as expected, whereas my data do not:

a <- 1980:2010
y <- sort( rnorm(30,2,.5) )
z <- matrix(nrow=30,ncol=31,1)

image(a, y, t(z), zlim = range(z), xlim = range(a), ylim = rev(range(y)), col = heat.colors(n=20))

using your data, I get a uniform image, what are your `xlim`, `zlim`, etc? — Rorschach, Aug 13 '15 at 21:09
just the range of x$t (1973 2012) and range of yvals (1.046901 3.796901) — Tiffany, Aug 13 '15 at 21:10
another note - when I plot the wtc object in the biwalevet package, I get the following error: Error in image.default(ix, iy, iz, xaxt = "n", yaxt = "n", xlab = "", :increasing 'x' and 'y' values expected — Tiffany, Aug 13 '15 at 21:21

score 0 · Answer 1 · answered Aug 14 '15 at 21:24

Turns out the biwavelet defaults for zlim were causing the issue in that the color scaling was altered along a gradient of very small scale - (e.g. 0.99-1), which was imposing different colors in the plot. I'm assuming that the precision displayed for the z values was less than what R was storing, accounting for the tiny discrepancies between values? The solution was to impose a wider range for the zlim (e.g. 0-1) which then forced the color gradient scheme to print appropriately (a single color).

R image function - error with color/value scaling

1 Answers1