rgl 2D surface plot of matrix not enough detail - r

I am trying to use rgl to surface plot this matrix "r":
library(rgl)
Banana Apple BlueBerry Kiwi Raisin Strawberry
Chicago 1.0000000 0.9972928 0.9947779 1.0623767 0.9976347 0.9993892
Wilmette 1.0016507 1.0000000 0.9976524 0.9863927 0.9985248 1.0016828
Winnetka 1.0040722 1.0025362 1.0000000 0.9886008 1.0016501 0.9955785
Glenview 0.9961316 1.0105463 1.0167024 1.0000000 1.0129399 1.0123440
Deerfield 1.0023308 1.0026052 0.9979093 0.9870921 1.0000000 1.0025606
Wheeling 1.0073697 0.9985745 1.0045129 0.9870925 1.0008054 1.0000000
rgl.surface(1:6 , 1:6 , r, color="red", back="lines")
Since the z-values are so close together in magnitude, the surface plot looks almost flat, even though there are subtle bumps in the data.
1) How do I make it so that I have a zoomed in version where I can see as much detail as possible?
2) Is there a way to show in different colors the data (faces) that have the biggest "slope", and so that the labels of the columns and rows of the matrix are preserved on the 3D surface (maybe just using the first three letters of the labels)? In other words, so I can see that the Kiwi in Chicago and the Kiwi in Wilmette causes the greatest min/max variation?

Something like this?
library(rgl)
library(colorRamps) # for matlab.like(...)
palette <- matlab.like(10) # palette of 10 colors
rlim <- range(r[!is.na(r)])
colors <- palette[9*(r-rlim[1])/diff(rlim) + 1]
open3d(scale=c(1/6,1/6,1/diff(range(r))))
surface3d(1:6 , 1:6 , r, color=colors, back="lines")
Part of your problem is that you were using rgl.surface(...) incorrectly. The second argument is the matrix of z-values. With surface3d(...) the arguments are x, y ,z in that order.
EDIT: Response to OP's comment.
Using your ex post facto dataset...
open3d(scale=c(1/6,1/6,1/diff(range(r))))
bbox3d(color="white")
surface3d(1:6 , 1:6 , r, color=colors, back="lines")
axis3d('x--',labels=rownames(r),tick=TRUE)
axis3d('y-+',labels=colnames(r),tick=TRUE)
axis3d('z--',tick=TRUE)

Related

How to visualize a distance matrix on the map by the thickness or color of the line connect the adjacent localities in R?

Suppose I have two datasets: (1) a data frame: coordinates of localities, each with ID; and (2) a linguistic distance matrix which reflects the linguistic distance between these localities.
# My data are similar to this structure
# dataframe
id <- c("A","B","C","D","E")
x_coor <- c(0.5,1,1,1.5,2)
y_coor <- c(5.5,3,7,6.5,5)
my.data <- data.frame(id = id, x_coor = x_coor, y_coor = y_coor)
# linguistic distance matrix
A B C D
B 308.298557
C 592.555483 284.256926
D 141.421356 449.719913 733.976839
E 591.141269 282.842712 1.414214 732.562625
Now, I want to visualize the linguistic distance between every two sites onto a map by the thickness or color of the line connect the adjacent localities in R.
Just like this:
enter image description here
My idea is to generate the delaunay triangulation by deldir or tripack package in R.
# generate delaunay triangulation
library(deldir)
de=deldir(my.data$x_coor,my.data$y_coor)
plot.deldir(de,wlines="triang",col='blue',wpoints = "real",cex = 0.1)
text(my.data$x_coor,my.data$y_coor,my.data$id)
this is the plot:
enter image description here
My question is how to reflect the linguistic distance by the thickness or color of the edges of triangles? Is there any other better method?
Thank you very much!
What you want to do in respect of the line widths can be done "fairly
easily" by the deldir package. You simply call plot.deldir() with the
appropriate value of "lw" (line width).
At the bottom of this answer is a demonstration script "demo.txt" which shows how to do this in the case of your example. In particular this script shows
how to obtain the appropriate value of lw from the "linguistic distance
matrix". I had to make some adjustments in the way this matrix was
presented. I.e. I had to convert it into a proper matrix.
I have rescaled the distances to lie between 0 and 10 to obtain the
corresponding values of the line widths. You might wish to rescale in a different manner.
In respect of colours, there are two issues:
(1) It is not at all clear how you would like to map the "linguistic
distances" to colours.
(2) Unfortunately the code for plot.deldir() is written in a very
kludgy way, whence the "col" argument to segments() cannot be
appropriately passed on in the same manner that the "lw" argument can.
(I wrote the plot.deldir() code a long while ago, when I knew far less about
R programming than I know now! :-))
I will adjust this code and submit a new version of deldir to CRAN
fairly soon.
#
# Demo script
#
# Present the linguistic distances in a useable way.
vldm <- c(308.298557,592.555483,284.256926,141.421356,449.719913,
733.976839,591.141269,282.842712,1.414214,732.562625)
ldm <- matrix(nrow=5,ncol=5)
ldm[row(ldm) > col(ldm)] <- vldm
ldm[row(ldm) <= col(ldm)] <- 0
ldm <- (ldm + t(ldm))/2
rownames(ldm) <- LETTERS[1:5]
colnames(ldm) <- LETTERS[1:5]
# Set up the example data. It makes life much simpler if
# you denote the "x" and "y" coordinates by "x" and "y"!!!
id <- c("A","B","C","D","E")
x_coor <- c(0.5,1,1,1.5,2)
y_coor <- c(5.5,3,7,6.5,5)
# Eschew nomenclature like "my.data". Such nomenclature
# is Micro$oft-ese and is an abomination!!!
demoDat <- data.frame(id = id, x = x_coor, y = y_coor)
# Form the triangulation/tessellation.
library(deldir)
dxy <- deldir(demoDat)
# Plot the triangulation with line widths proportional
# to "linguistic distances". Note that plot.deldir() is
# a *method* for plot, so you do not have to (and shouldn't)
# type the ".deldir" in the plotting command.
plot(dxy,col=0) # This, and plotting with "add=TRUE" below, is
# a kludge to dodge around spurious warnings.
ind <- as.matrix(dxy$delsgs[,c("ind1","ind2")])
lwv <- ldm[ind]
lwv <- 10*lwv/max(lwv)
plot(dxy,wlines="triang",col='grey',wpoints="none",
lw=10*lwv/max(lwv),add=TRUE)
with(demoDat,text(x,y,id,col="red",cex=1.5))

Selecting overlapping points on a plot

I have two matrices which are built as follows
x1=cbind(V1,V2,ID)
X2=cbind(V1,V2,ID)
X3=rbind(X1,X2)
ID takes only the values "red" and "blue"
when I plot X1 and X2 I have the following plot
I want to select the data points which are within 1 unit distance (euclidian distance) basically filtering only the red points which are overlapping or quasi-overlapping a blue point or vice versa.
Red overlapping red and blue overlapping blue is not interesting for me.
Thanks a lot for your assistance.
You definitely need to provide a reproducible example for this one to get the best answer; however, I think below script will serve the purpose:
library(spatstat)
# setting seeds
set.seed(222)
# two different point patterns
X <- runifpoint(15)
Y <- runifpoint(20)
plot(X, pch=19, main="")
plot(Y, col="red", pch=19, add=T)
#you can get both which and dist from nncross
#N.which <- nncross(X,Y, k=1:20, what="which")
#N.dist <- nncross(X,Y, k=1:20, what="dist")
out <- subset(X, nncross(X,Y, k=1:20, what="dist") < 0.1) # you may change 0.1
plot(out, col="blue", pch=19, add=T)
For the above plot, black points represent X and red points represent Y. Blue are intersecting points which are within 0.1 unit distance. This distance can be further modified. For more detaild please see spatstat to compute distances between two different datasets using nncross.

Thin Plate Spline for 3D surface prediction in R

I tried this answer
get a surface plot in R
but it hasn't really helped.
I would like to perform a TPS (using Tps from Fields{}) on an XYZ dataframe where xy are co-orinates and z is a thickness. Then I would like to visualise the plot firstly before TPS and then after TPS..? Is this possible.
Then I would like to extract predicted thicknesses for a given set of new xy co-ordinates..?
Please let me know if this is possible
My Dataframe looks like this, dataframe is called LSP:
time PART MEAS PARTSUB XLOC YLOC
xxxx 1 1.956 a -3465 -94350
xxxx 1 1.962 a -3465 -53850
xxxx 1 1.951 a 50435 -40350
xxxx 1 1.958 a -57365 -40350
So I tried this:
LSP.spline <- Tps(LSP[,5:6], LSP$MEAS)
out.p <- predict.surface(LSP.spline, xy = c(1,2))
plot.surface(out.p, type="p")
But out.p is just NULL..?
so attempting the plot gives me:
Error in nrow(z) : argument "z" is missing, with no default
Any help is appreciated.
Paul.
predict.surface is now an obsolete / deprecated function. Use predictSurface instead.
fit<- Tps( BD[,1:4], BD$lnya) # fit surface to data
# evaluate fitted surface for first two
# variables holding other two fixed at median values
out.p<- predictSurface(fit)
surface(out.p, type="C")
Thanks for that - how about my second question....how can I extract predicted surface thickness values for a given set of XY locations..?
Use predict function. Have a read on ?predict.Tps. For the above example, let's say we want to predict at the first 4 locations in BD[, 1:4], we can do
predict(fit, x = BD[1:4, 1:4])
# [,1]
#[1,] 11.804124
#[2,] 11.804124
#[3,] 8.069056
#[4,] 9.501551
In general, pass x a two-column matrix.

Group bar plot with error bars and spit y axis

I would like to draw a group bar graph with error bars and split y axis to show both smaller and larger values in same plot? (as shown in my data sample number 1 has small values compare to other samples, therefore, I want to make a gap on y axis in-between 10-200)
Here is my data,
sample mean part sd
1 4.3161 G 1.2209
1 2.3157 F 1.7011
1 1.7446 R 1.1618
2 1949.13 G 873.42
2 195.07 F 47.82
2 450.88 R 140.31
3 2002.98 G 367.92
3 293.45 F 59.01
3 681.99 R 168.03
4 2717.85 G 1106.07
4 432.83 F 118.02
4 790.97 R 232.62
You can do anything you want with primitive graphic elements. For this reason, I always prefer to design my own plots with just the base R plotting functions, particularly points(), segments(), lines(), abline(), rect(), polygon(), text(), and mtext(). You can easily create curves (e.g. for circles) and more complex shapes using segments() and lines() across granular coordinate vectors that you define yourself. For example, see Plot angle between vectors. This provides much more control over the plot elements you create, however, it often takes more work and careful coding than more prepackaged solutions, so it's a tradeoff.
Data
First, here's your data in runnable form:
df <- data.frame(
sample=c(1,1,1,2,2,2,3,3,3,4,4,4),
mean=c(4.3161,2.3157,1.7446,1949.13,195.07,450.88,2002.98,293.45,681.99,2717.85,432.83,790.97),
part=c('G','F','R','G','F','R','G','F','R','G','F','R'),
sd=c(1.2209,1.7011,1.1618,873.42,47.82,140.31,367.92,59.01,168.03,1106.07,118.02,232.62),
stringsAsFactors=F
);
df;
## sample mean part sd
## 1 1 4.3161 G 1.2209
## 2 1 2.3157 F 1.7011
## 3 1 1.7446 R 1.1618
## 4 2 1949.1300 G 873.4200
## 5 2 195.0700 F 47.8200
## 6 2 450.8800 R 140.3100
## 7 3 2002.9800 G 367.9200
## 8 3 293.4500 F 59.0100
## 9 3 681.9900 R 168.0300
## 10 4 2717.8500 G 1106.0700
## 11 4 432.8300 F 118.0200
## 12 4 790.9700 R 232.6200
OP ggplot
Now, for reference, here's a screenshot of the plot that results from the ggplot code you pasted into your comment:
library(ggplot2);
ggplot(df,aes(x=as.factor(sample),y=mean,fill=part)) +
geom_bar(position=position_dodge(),stat='identity',colour='black') +
geom_errorbar(aes(ymin=mean-sd,ymax=mean+sd),width=.2,position=position_dodge(.9));
Linear Single
Also for reference, here's how you can produce a similar grouped bar plot using base R barplot() and legend(). I've added the error bars with custom calls to segments() and points():
## reshape to wide matrices
dfw <- reshape(df,dir='w',idvar='part',timevar='sample');
dfw.mean <- as.matrix(dfw[grep(perl=T,'^mean\\.',names(dfw))]);
dfw.sd <- as.matrix(dfw[grep(perl=T,'^sd\\.',names(dfw))]);
rownames(dfw.mean) <- rownames(dfw.sd) <- dfw$part;
colnames(dfw.mean) <- colnames(dfw.sd) <- unique(df$sample);
## plot precomputations
ylim <- c(0,4000);
yticks <- seq(ylim[1L],ylim[2L],100);
xcenters <- (col(dfw.sd)-1L)*(nrow(dfw.sd)+1L)+row(dfw.sd)+0.5;
partColors <- c(G='green3',F='indianred1',R='dodgerblue');
errColors <- c(G='darkgreen',F='darkred',R='darkblue');
## plot
par(xaxs='i',yaxs='i');
barplot(dfw.mean,beside=T,col=partColors,ylim=ylim,xlab='sample',ylab='mean',axes=F);
segments(xcenters,dfw.mean-dfw.sd,y1=dfw.mean+dfw.sd,lwd=2,col=errColors);
points(rep(xcenters,2L),c(dfw.mean-dfw.sd,dfw.mean+dfw.sd),pch=19,col=errColors);
axis(1L,par('usr')[1:2],F,pos=0,tck=0);
axis(2L,yticks,las=1L,cex.axis=0.7);
legend(2,3800,dfw$part,partColors,title=expression(bold('part')),cex=0.7,title.adj=0.5[2:1]);
The issue is plain to see. There's nuance to some of the data (the sample 1 means and variability) that is not well represented in the plot.
Logarithmic
There are two standard options for dealing with this problem. One is to use a logarithmic scale. You can do this with the log='y' argument to the barplot() function. It's also good to override the default y-axis tick selection, since the default base R ticks tend to be a little light on density and short on range. (That's actually true in general, for most base R plot types; I make custom calls to axis() for all the plots I produce in this answer.)
## plot precomputations
ylim <- c(0.1,4100); ## lower limit must be > 0 for log plot
yticks <- rep(10^seq(floor(log10(ylim[1L])),ceiling(log10(ylim[2L])),1),each=9L)*1:9;
xcenters <- (col(dfw.sd)-1L)*(nrow(dfw.sd)+1L)+row(dfw.sd)+0.5;
partColors <- c(G='green3',F='indianred1',R='dodgerblue');
errColors <- c(G='darkgreen',F='darkred',R='darkblue');
## plot
par(xaxs='i',yaxs='i');
barplot(log='y',dfw.mean,beside=T,col=partColors,ylim=ylim,xlab='sample',ylab='mean',axes=F);
segments(xcenters,dfw.mean-dfw.sd,y1=dfw.mean+dfw.sd,lwd=2,col=errColors);
points(rep(xcenters,2L),c(dfw.mean-dfw.sd,dfw.mean+dfw.sd),pch=19,col=errColors);
axis(1L,par('usr')[1:2],F,pos=0,tck=0);
axis(2L,yticks,yticks,las=1L,cex.axis=0.6);
legend(2,3000,dfw$part,partColors,title=expression(bold('part')),cex=0.7,title.adj=0.5[2:1]);
Right away we see the issue with sample 1 is fixed. But we've introduced a new issue: we've lost precision in the rest of the data. In other words, the nuance that exists in the rest of the data is less visually pronounced. This is an unavoidable result of the "zoom-out" effect of changing from linear to logarithmic axes. You would incur the same loss of precision if you used a linear plot but with too large a y-axis, which is why it's always expected that axes are fitted as closely as possible to the data. This also serves as an indication that a logarithmic y-axis may not be the correct solution for your data. Logarithmic axes are generally advised when the underlying data reflects logarithmic phenomena; that it ranges over several orders of magnitude. In your data, only sample 1 sits in a different order of magnitude from the remaining data; the rest are concentrated in the same order of magnitude, and are thus not best represented with a logarithmic y-axis.
Linear Multiple
The second option is to create separate plots with completely different y-axis scaling. It should be noted that ggplot faceting is essentially the creation of separate plots. Also, you could create multifigure plots with base R, but I've usually found that that's more trouble than it's worth. It's usually easier to just generate each plot individually, and then lay them out next to each other with publishing or word processing software.
There are different ways of customizing this approach, such as whether you combine the axis labels, where you place the legend, how you size and arrange the different plots relative to each other, etc. Here's one way of doing it:
##--------------------------------------
## plot 1 -- high values
##--------------------------------------
dfw.mean1 <- dfw.mean[,-1L];
dfw.sd1 <- dfw.sd[,-1L];
## plot precomputations
ylim <- c(0,4000);
yticks <- seq(ylim[1L],ylim[2L],100);
xcenters <- (col(dfw.sd1)-1L)*(nrow(dfw.sd1)+1L)+row(dfw.sd1)+0.5;
partColors <- c(G='green3',F='indianred1',R='dodgerblue');
errColors <- c(G='darkgreen',F='darkred',R='darkblue');
par(xaxs='i',yaxs='i');
barplot(dfw.mean1,beside=T,col=partColors,ylim=ylim,xlab='sample',ylab='mean',axes=F);
segments(xcenters,dfw.mean1-dfw.sd1,y1=dfw.mean1+dfw.sd1,lwd=2,col=errColors);
points(rep(xcenters,2L),c(dfw.mean1-dfw.sd1,dfw.mean1+dfw.sd1),pch=19,col=errColors);
axis(1L,par('usr')[1:2],F,pos=0,tck=0);
axis(2L,yticks,las=1L,cex.axis=0.7);
legend(2,3800,dfw$part,partColors,title=expression(bold('part')),cex=0.7,title.adj=0.5[2:1]);
##--------------------------------------
## plot 2 -- low values
##--------------------------------------
dfw.mean2 <- dfw.mean[,1L,drop=F];
dfw.sd2 <- dfw.sd[,1L,drop=F];
## plot precomputations
ylim <- c(0,6);
yticks <- seq(ylim[1L],ylim[2L],0.5);
xcenters <- (col(dfw.sd2)-1L)*(nrow(dfw.sd2)+1L)+row(dfw.sd2)+0.5;
partColors <- c(G='green3',F='indianred1',R='dodgerblue');
errColors <- c(G='darkgreen',F='darkred',R='darkblue');
par(xaxs='i',yaxs='i');
barplot(dfw.mean2,beside=T,col=partColors,ylim=ylim,xlab='sample',ylab='mean',axes=F);
segments(xcenters,dfw.mean2-dfw.sd2,y1=dfw.mean2+dfw.sd2,lwd=2,col=errColors);
points(rep(xcenters,2L),c(dfw.mean2-dfw.sd2,dfw.mean2+dfw.sd2),pch=19,col=errColors);
axis(1L,par('usr')[1:2],F,pos=0,tck=0);
axis(2L,yticks,las=1L,cex.axis=0.7);
This solves both problems (small-value visibility and large-value precision). But it also distorts the relative magnitude of samples 2-4 vs. sample 1. In other words, the sample 1 data has been "scaled up" relative to samples 2-4, and the reader must make a conscious effort to read the axes and digest the differing scales in order to properly understand the plots.
The lesson here is that there's no perfect solution. Every approach has its own pros and cons, its own tradeoffs.
Gapped
In your question, you indicate you want to add a gap across the y range 10:200. On the surface, this sounds like a reasonable solution for raising the visibility of the sample 1 data. However, the magnitude of that 190 unit range is dwarfed by the range of the remainder of the plot, so it ends up having a negligible effect on sample 1 visibility.
In order to demonstrate this I'm going to use some code I've written which can be used to transform input coordinates to a new data domain which allows for inconsistent scaling of different segments of the axis. Theoretically you could use it for both x and y axes, but I've only ever used it for the y-axis.
A few warnings: This introduces some significant complexity, and decouples the graphics engine's idea of the y-axis scale from the real data. More specifically, it maps all coordinates to the range [0,1] based on their cumulative position within the sequence of segments.
At this point, I'm also going to abandon barplot() in favor of drawing the bars manually, using calls to rect(). Technically, it would be possible to use barplot() with my segmentation code, but as I said earlier, I prefer to design my own plots from scratch with primitive graphic elements. This also allows for more precise control over all aspects of the plot.
Here's the code and plot, I'll attempt to give a better explanation of it afterward:
dataCoordToPlot <- function(data,seg) {
## data -- double vector of data-world coordinates.
## seg -- list of two components: (1) mark, giving the boundaries between all segments, and (2) scale, giving the relative scale of each segment. Thus, scale must be one element shorter than mark.
data <- as.double(data);
seg <- as.list(seg);
seg$mark <- as.double(seg$mark);
seg$scale <- as.double(seg$scale);
if (length(seg$scale) != length(seg$mark)-1L) stop('seg$scale must be one element shorter than seg$mark.');
scaleNorm <- seg$scale/sum(seg$scale);
cumScale <- c(0,cumsum(scaleNorm));
int <- findInterval(data,seg$mark,rightmost.closed=T);
int[int%in%c(0L,length(seg$mark))] <- NA; ## handle values outside outer segments; will propagate NA to returned vector
(data - seg$mark[int])/(seg$mark[int+1L] - seg$mark[int])*scaleNorm[int] + cumScale[int];
}; ## end dataCoordToPlot()
## y dimension segmentation
ymax <- 4000;
yseg <- list();
yseg$mark <- c(0,10,140,ymax);
yseg$scale <- diff(yseg$mark);
yseg$scale[2L] <- 30;
yseg$jump <- c(F,T,F);
## plot precomputations
xcenters <- seq(0.5,len=length(unique(df$sample)));
xlim <- range(xcenters)+c(-0.5,0.5);
ylim <- range(yseg$mark);
yinc <- 100;
yticks.inc <- seq(ylim[1L],ylim[2L],yinc);
yticks.inc <- yticks.inc[!yseg$jump[findInterval(yticks.inc,yseg$mark,rightmost.closed=T)]];
yticks.jump <- setdiff(yseg$mark,yticks.inc);
yticks.all <- sort(c(yticks.inc,yticks.jump));
## plot
## define as reusable function for subsequent examples
custom.barplot <- function() {
par(xaxs='i',yaxs='i');
plot(NA,xlim=xlim,ylim=dataCoordToPlot(ylim,yseg),axes=F,ann=F);
abline(h=dataCoordToPlot(yticks.all,yseg),col='lightgrey');
axis(1L,seq(xlim[1L],xlim[2L]),NA,tck=0);
axis(1L,xcenters,unique(df$sample));
axis(2L,dataCoordToPlot(yticks.inc,yseg),yticks.inc,las=1,cex.axis=0.7);
axis(2L,dataCoordToPlot(yticks.jump,yseg),yticks.jump,las=1,tck=-0.008,hadj=0.1,cex.axis=0.5);
mtext('sample',1L,2L);
mtext('mean',2L,3L);
xgroupRatio <- 0.8;
xbarRatio <- 0.9;
partColors <- c(G='green3',F='indianred1',R='dodgerblue');
partsCanon <- unique(df$part);
errColors <- c(G='darkgreen',F='darkred',R='darkblue');
for (sampleIndex in seq_along(unique(df$sample))) {
xc <- xcenters[sampleIndex];
sample <- unique(df$sample)[sampleIndex];
dfs <- df[df$sample==sample,];
parts <- unique(dfs$part);
parts <- parts[order(match(parts,partsCanon))];
barWidth <- xgroupRatio*xbarRatio/length(parts);
gapWidth <- xgroupRatio*(1-xbarRatio)/(length(parts)-1L);
xstarts <- xc - xgroupRatio/2 + (match(dfs$part,parts)-1L)*(barWidth+gapWidth);
rect(xstarts,0,xstarts+barWidth,dataCoordToPlot(dfs$mean,yseg),col=partColors[dfs$part]);
barCenters <- xstarts+barWidth/2;
segments(barCenters,dataCoordToPlot(dfs$mean + dfs$sd,yseg),y1=dataCoordToPlot(dfs$mean - dfs$sd,yseg),lwd=2,col=errColors);
points(rep(barCenters,2L),dataCoordToPlot(c(dfs$mean-dfs$sd,dfs$mean+dfs$sd),yseg),pch=19,col=errColors);
}; ## end for
## draw zig-zag cutaway graphic in jump segments
zigCount <- 30L;
jumpIndexes <- which(yseg$jump);
for (jumpIndex in jumpIndexes) {
if (yseg$scale[jumpIndex] == 0) next;
jumpStart <- yseg$mark[jumpIndex];
jumpEnd <- yseg$mark[jumpIndex+1L];
lines(seq(xlim[1L],xlim[2L],len=zigCount*2L+1L),dataCoordToPlot(c(rep(c(jumpStart,jumpEnd),zigCount),jumpStart),yseg));
}; ## end for
legend(0.2,dataCoordToPlot(3800,yseg),partsCanon,partColors,title=expression(bold('part')),cex=0.7,title.adj=c(NA,0.5));
}; ## end custom.barplot()
custom.barplot();
The key function is dataCoordToPlot(). That stands for "data coordinates to plot coordinates", where "plot coordinates" refers to the [0,1] normalized domain.
The seg argument defines the segmentation of the axis and the scaling of each segment. Its mark component specifies the boundaries of each segment, and its scale component gives the scale factor for each segment. n segments must have n+1 boundaries to fully define where each segment begins and ends, thus mark must be one element longer than scale.
Before being used, the scale vector is normalized within the function to sum to 1, so the absolute magnitudes of the scale values don't matter; it's their relative values that matter.
The algorithm is to find each coordinate's containing segment, find the accumulative distance within the segment reached by the coordinate accounting for the segment's relative scale, and then add to that the cumulative distance reached by all prior segments.
Using this design, it is possible to take any range of coordinates along the axis dimension and scale them up or down relative to the other segments. An instantaneous gap across a range could be achieved with a scale of zero. Alternatively, you can simply scale down the range so that it has some thickness, but contributes little to the progression of the dimension. In the above plot, I use the latter for the gap, mainly so that I can use the small thickness to add a zigzag aesthetic which visually indicates the presence of the gap.
Also, I should note that I used 10:140 instead of 10:200 for the gap. This is because the sample 2 F part error bar extends down to 147.25 (195.07 - 47.82). The difference is negligible.
As you can see, the result looks basically identical to the Linear Single plot. The gap is not significant enough to raise the visibility of the sample 1 data.
Distorted with Gap
Just to throw some more possibilities into mix, now venturing into very non-standard and probably questionable waters, we can use the segmentation transformation to scale up the sample 1 order of magnitude, thereby making it much more visible while still remaining within the single plot, directly alongside samples 2-4.
For this example, I preserve the gap from 10:140 so you can see how it looks when not lying prostrate near the baseline.
## y dimension segmentation
ymax <- 4000;
yseg <- list();
yseg$mark <- c(0,10,140,ymax);
yseg$scale <- c(24,1,75);
yseg$jump <- c(F,T,F);
## plot precomputations
xcenters <- seq(0.5,len=length(unique(df$sample)));
xlim <- range(xcenters)+c(-0.5,0.5);
ylim <- range(yseg$mark);
yinc1 <- 1;
yinc2 <- 100;
yticks.inc1 <- seq(ceiling(yseg$mark[1L]/yinc1)*yinc1,yseg$mark[2L],yinc1);
yticks.inc2 <- seq(ceiling(yseg$mark[3L]/yinc2)*yinc2,yseg$mark[4L],yinc2);
yticks.inc <- c(yticks.inc1,yticks.inc2);
yticks.jump <- setdiff(yseg$mark,yticks.inc);
yticks.all <- sort(c(yticks.inc,yticks.jump));
## plot
custom.barplot();
Distorted without Gap
Finally, just to clarify that gaps are not necessary for inconsistent scaling between segments, here's the same plot but without the gap:
## y dimension segmentation
ymax <- 4000;
yseg <- list();
yseg$mark <- c(0,10,ymax);
yseg$scale <- c(25,75);
yseg$jump <- c(F,F);
## plot precomputations
xcenters <- seq(0.5,len=length(unique(df$sample)));
xlim <- range(xcenters)+c(-0.5,0.5);
ylim <- range(yseg$mark);
yinc1 <- 1;
yinc2 <- 100;
yticks.inc1 <- seq(ceiling(yseg$mark[1L]/yinc1)*yinc1,yseg$mark[2L],yinc1);
yticks.inc2 <- seq(ceiling(yseg$mark[2L]/yinc2)*yinc2,yseg$mark[3L],yinc2);
yticks.inc <- c(yticks.inc1,yticks.inc2);
yticks.jump <- setdiff(yseg$mark,yticks.inc);
yticks.all <- sort(c(yticks.inc,yticks.jump));
## plot
custom.barplot();
In principle, there's really no difference between the Linear Multiple solution and the Distorted solutions. Both involve visual distortion of competing orders of magnitude. Linear Multiple simply separates the different orders of magnitude into separate plots, while the Distorted solutions combine them into the same plot.
Probably the best argument in favor of using Linear Multiple is that if you use Distorted you'll probably be crucified by a large mob of data scientists, since that is a very non-standard way of plotting data. On the other hand, one could argue that the Distorted approach is more concise and helps to represent the relative positions of each data point along the number line. The choice is yours.
What you want to plot is a discontinuous y axis.
This issue was covered before in this post and seems not to be possible in ggplot2.
The answers to the mentioned post suggest faceting, log scaled y axis and separate plots to solve your problem.
Please find the reasons detailed by Hadley Wickham here, who thinks that a broken y axis could be "visually distorting".

Graph to compare two matrices in R

I have two matrices (of approximately 300 x 100) and I would like to plot a graph to see the parts of the first one that are higher than those of the second.
I can do, for instance:
# Calculate the matrices and put them into m1 and m2
# Note that the values are between -1 and 1
par(mfrow=c(1,3))
image(m1, zlim=c(-1,1))
image(m2, zlim=c(-1,1))
image(m1-m2, zlim=c(0,1))
This will plot only the desired regions in the 3rd plot but I would like to do something a bit different, like putting a line around those areas over the first plot in order to highlight them directly there.
Any idea how I can do that?
Thank you
nico
How about:
par(mfrow = c(1, 3))
image(m1, zlim = c(-1, 1))
contour(m1 - m2, add = TRUE)
image(m2, zlim = c(-1, 1))
contour(m1 - m2, add = TRUE)
image(m1 - m2, zlim = c(0, 1))
contour(m1 - m2, add = TRUE)
This adds a contour map around the regions. Sort of puts rings around the areas of the 3rd plot (might want to fiddle with the (n)levels of the contours to get fewer 'circles').
Another way of doing your third image might be:
image(m1>m2)
this produces a matrix of TRUE/FALSE values which gets imaged as 0/1, so you have a two-colour image. Still not sure about your 'putting a line around' thing though...
Here's some code I wrote to do something similar. I wanted to highlight contiguous regions above a 0.95 threshold by drawing a box round them, so I got all the grid squares above 0.95 and did a clustering on them. Then do a bit of fiddling with the clustering output to get the rectangle coordinates of the regions:
computeHotspots = function(xyz, thresh, minsize=1, margin=1){
### given a list(x,y,z), return a data frame where each row
### is a (xmin,xmax,ymin,ymax) of bounding box of a contiguous area
### over the given threshhold.
### or approximately. lets use the clustering tools in R...
overs <- which(xyz$z>thresh,arr.ind=T)
if(length(overs)==0){
## found no hotspots
return(NULL)
}
if(length(overs)==2){
## found one hotspot
xRange <- cbind(xyz$x[overs[,1]],xyz$x[overs[,1]])
yRange <- cbind(xyz$y[overs[,2]],xyz$y[overs[,2]])
}else{
oTree <- hclust(dist(overs),method="single")
oCut <- cutree(oTree,h=10)
oXYc <- data.frame(x=xyz$x[overs[,1]],y=xyz$y[overs[,2]],oCut)
xRange <- do.call("rbind",tapply(oXYc[,1],oCut,range))
yRange <- do.call("rbind",tapply(oXYc[,2],oCut,range))
}
### add user-margins
xRange[,1] <- xRange[,1]-margin
xRange[,2] <- xRange[,2]+margin
yRange[,1] <- yRange[,1]-margin
yRange[,2] <- yRange[,2]+margin
## put it all together
xr <- apply(xRange,1,diff)
xm <- apply(xRange,1,mean)
xRange[xr<minsize,1] <- xm[xr<minsize]-(minsize/2)
xRange[xr<minsize,2] <- xm[xr<minsize]+(minsize/2)
yr <- apply(yRange,1,diff)
ym <- apply(yRange,1,mean)
yRange[yr<minsize,1] <- ym[yr<minsize]-(minsize/2)
yRange[yr<minsize,2] <- ym[yr<minsize]+(minsize/2)
cbind(xRange,yRange)
}
Test code:
x=1:23
y=7:34
m1=list(x=x,y=y,z=outer(x,y,function(x,y){sin(x/3)*cos(y/3)}))
image(m1)
hs = computeHotspots(m1,0.95)
That should give you a matrix of rectangle coordinates:
> hs
[,1] [,2] [,3] [,4]
1 13 15 8 11
2 3 6 17 20
3 22 24 18 20
4 13 16 27 30
Now you can draw them over the image with rect:
image(m1)
rect(hs[,1],hs[,3],hs[,2],hs[,4])
and to show they are where they should be:
image(list(x=m1$x,y=m1$y,z=m1$z>0.95))
rect(hs[,1],hs[,3],hs[,2],hs[,4])
You could of course adapt this to draw circles, but more complex shapes would be tricky. It works best when the regions of interest are fairly compact.
Barry

Resources