I want to set the starting point of my graph (both in the x- and y-axis) to be zero. I have used the xlim() and ylim() functions to no avail. I have attached my code below.
setwd ("D:/Rcode/Assignment_2") #setting up the working directory
LightGrowth1 <- read.csv ("LightGrowth-1.csv") #reading the file and attaching it to a dataframe
Light <- LightGrowth1$light #attach our light values to a vector in R
Growth <- LightGrowth1$growth #attach our growth values to a vector in R
Labels <- c("Light", "Growth") #create a vector using the labels
plot (Light, Growth, xlab = "Amount of Light (units)", ylab = "Plant Growth (units)",
pch = 16, col= "firebrick", xlim = c(0, max (Light)), ylim = c (0, max (Growth)),
main = "Plant Growth vs Amount of Light"
)
This is what my plot currently looks like:
Your axes do start at 0, and xlim and ylim are what you need to change that.
What I think you are referring to is the spacing between the axes and the plot.
There are two extra parameters that allow you to change that, xaxs and yaxs.
plot (x, y, xlim=c(0,10), ylim=c(0,10),
xaxs="i", yaxs="i")
This should do the trick
See the par help for more info.
I want to set the starting point of my graph (both in the x- and y-axis) to be zero. I have used the xlim() and ylim() functions to no avail. I have attached my code below.
setwd ("D:/Rcode/Assignment_2") #setting up the working directory
LightGrowth1 <- read.csv ("LightGrowth-1.csv") #reading the file and attaching it to a dataframe
Light <- LightGrowth1$light #attach our light values to a vector in R
Growth <- LightGrowth1$growth #attach our growth values to a vector in R
Labels <- c("Light", "Growth") #create a vector using the labels
plot (Light, Growth, xlab = "Amount of Light (units)", ylab = "Plant Growth (units)",
pch = 16, col= "firebrick", xlim = c(0, max (Light)), ylim = c (0, max (Growth)),
main = "Plant Growth vs Amount of Light"
)
This is what my plot currently looks like:
Your axes do start at 0, and xlim and ylim are what you need to change that.
What I think you are referring to is the spacing between the axes and the plot.
There are two extra parameters that allow you to change that, xaxs and yaxs.
plot (x, y, xlim=c(0,10), ylim=c(0,10),
xaxs="i", yaxs="i")
This should do the trick
See the par help for more info.
I've successfully produced NMDS plots (monoMDS, bray-curtis, 3 dimensions, local model). Each point represents an animal and their diet composition.
I have two questions:
(1) how do I change the symbology of points to show 2 levels (a or j) within 1 column (Life stage) on the NMDS plot?!
(2) How should I show 3D NMDS, I can't get the 3D orgl- functions to work on the 3D plot. Should I just make a few plots showing different dimensions in 2D? Looking for thoughtful ideas.
The code used:
plot((BC.NMDS.length.corr), choices = c(1, 2), type = "points",
xlim = c(-2.0, 2.0),las = 1, ylim = c(-1, 1),
xlab = "NMDS Axis 1", ylab = "NMDS Axis 2",mgp = c(3.25, 1, 0),
cex.lab = 1.35, cex.axis = 1.25)
with(DATA,
points(BC.NMDS.length.corr, Class, draw = "points",col = "gray0",
show.groups = "Adult",label = TRUE, lty = 1, lwd = 2))
Using an example of what you want with the default example of the package:
# Load library
library(vegan)
# Load data
data(dune)
# Compute the distance
dis <- vegdist(dune)
Specify if you want a 3D plot, the representation of the three dimensions
# Run monoMDS
m <- monoMDS(dis, model = "loc", k=3)
# The 3D representation
plot(m)
# Load library for 3D representation
library(scatterplot3d)
Coordinates are in m$points; each column referring to each dimension.
# Graphical representation
scatterplot3d(x=m$points[,1], y=m$points[,2], z=m$points[,3])
Additionally, if you want to colour the plots depending on a factor, you can specify color=A, where A is a numeric value where groups are codified.
Given such data:
#Cutpoint SN (1-PPV)
5 0.56 0.01
7 0.78 0.19
9 0.91 0.58
How can I plot ROC curve with R that produce similar result like the
attached ?
I know ROCR package but it doesn't take such input.
If you just want to create the plot (without that silly interpolation spline between points) then just plot the data you give in the standard way, prepending a point at (0,0) and appending one at (1,1) to give the end points of the curve.
## your data with different labels
dat <- data.frame(cutpoint = c(5, 7, 9),
TPR = c(0.56, 0.78, 0.91),
FPR = c(0.01, 0.19, 0.58))
## plot version 1
op <- par(xaxs = "i", yaxs = "i")
plot(TPR ~ FPR, data = dat, xlim = c(0,1), ylim = c(0,1), type = "n")
with(dat, lines(c(0, FPR, 1), c(0, TPR, 1), type = "o", pch = 25, bg = "black"))
text(TPR ~ FPR, data = dat, pos = 3, labels = dat$cutpoint)
abline(0, 1)
par(op)
To explain the code: The first plot() call sets up the plotting region, without doing an plotting at all. Note that I force the plot to cover the range (0,1) in both axes. The par() call tells R to plot axes that cover the range of the data - the default extends them by 4 percent of the range on each axis.
The next line, with(dat, lines(....)) draws the ROC curve and here we prepend and append the points at (0,0) and (1,1) to give the full curve. Here I use type = "o" to give both points and lines overplotted, the points are represented by character 25 which allows it to be filled with a colour, here black.
Then I add labels to the points using text(....); the pos argument is used to position the label away from the actual plotting coordinates. I take the labels from the cutpoint object in the data frame.
The abline() call draws the 1:1 line (here the 0, and 1 mean an intercept of 0 and a slope of 1 respectively.
The final line resets the plotting parameters to the defaults we saved in op prior to plotting (in the first line).
The resulting plot looks like this:
It isn't an exact facsimile and I prefer the plot using the default for the axis ranges(adding 4 percent):
plot(TPR ~ FPR, data = dat, xlim = c(0,1), ylim = c(0,1), type = "n")
with(dat, lines(c(0, FPR, 1), c(0, TPR, 1), type = "o", pch = 25, bg = "black"))
text(TPR ~ FPR, data = dat, pos = 3, labels = dat$cutpoint)
abline(0, 1)
Again, not a true facsimile but close.
Is there a way to ensure that the box around a plot matches the raster extents exactly? In the following there is a gap above and below or to the left and right of the raster depending on the device proportions:
require(raster)
r = raster()
r[]= 1
plot(r, xlim=c(xmin(r), xmax(r)), ylim=c(ymin(r), ymax(r)))
One element of the problem with raster objects is that asp=1 to ensure proper display. The following basic scatterplot has the same issue when asp=1:
plot(c(1:10), c(1:10), asp=1)
Try vectorplot(r) from the rasterVis package to see what I want the axes to look like.
EDIT:
Solutions need to play nice with SpatialPoints overlays, not showing points outside the specified raster limits:
require(raster)
require(maptools)
# Raster
r = raster()
r[]= 1
# Spatial points
x = c(-100, 0, 100)
y = c(100, 0, 100)
points = SpatialPoints(data.frame(x,y))
plot(r, xlim=c(xmin(r), xmax(r)), ylim=c(ymin(r), ymax(r)))
plot(points, add=T)
You'd probably do best to go with one of the lattice-based functions for plotting spatial raster objects provided by the raster and rasterVis packages. You discovered one of them in vectorplot(), but spplot() or levelplot() better match your needs in this case.
(The base graphics-based plot() method for "RasterLayer" objects just doesn't allow any easy way for you to set axes with the appropriate aspect ratio. For anyone interested, I go into more detail about why that's so in a section at the bottom of the post.)
As an example of the kind of plot that levelplot() produces:
require(raster)
require(rasterVis)
## Create a raster and a SpatialPoints object.
r <- raster()
r[] <- 1:ncell(r)
SP <- spsample(Spatial(bbox=bbox(r)), 10, type="random")
## Then plot them
levelplot(r, col.regions = rev(terrain.colors(255)), cuts=254, margin=FALSE) +
layer(sp.points(SP, col = "red"))
## Or use this, which produces the same plot.
# spplot(r, scales = list(draw=TRUE),
# col.regions = rev(terrain.colors(255)), cuts=254) +
# layer(sp.points(SP, col = "red"))
Either of these methods may still plot some portion of the symbol representing points that fall just outside of the plotted raster. If you want to avoid that possibility, you can just subset your SpatialPoints object to remove any points falling outside of the raster. Here's a simple function that'll do that for you:
## A function to test whether points fall within a raster's extent
inExtent <- function(SP_obj, r_obj) {
crds <- SP_obj#coord
ext <- extent(r_obj)
crds[,1] >= ext#xmin & crds[,1] <= ext#xmax &
crds[,2] >= ext#ymin & crds[,2] <= ext#ymax
}
## Remove any points in SP that don't fall within the extent of the raster 'r'
SP <- SP[inExtent(SP, r), ]
Additional weedy detail about why it's hard to make plot(r) produce snugly fitting axes
When plot is called on an object of type raster, the raster data is (ultimately) plotted using either rasterImage() or image(). Which path is followed depends on: (a) the type of device being plotted to; and (b) the value of the useRaster argument in the original plot() call.
In either case, the plotting region is set up in a way which produces axes that fill the plotting region, rather than in a way that gives them the appropriate aspect ratio.
Below, I show the chain of functions that's called on the way to this step, as well as the call that ultimately sets up the plotting region. In both cases, there appears to be no simple way to alter both the extent and the aspect ratio of the axes that are plotted.
useRaster=TRUE
## Chain of functions dispatched by `plot(r, useRaster=TRUE)`
getMethod("plot", c("RasterLayer", "missing"))
raster:::.plotraster2
raster:::.rasterImagePlot
## Call within .rasterImagePlot() that sets up the plotting region
plot(NA, NA, xlim = e[1:2], ylim = e[3:4], type = "n",
, xaxs = "i", yaxs = "i", asp = asp, ...)
## Example showing why the above call produces the 'wrong' y-axis limits
plot(c(-180,180), c(-90,90),
xlim = c(-180,180), ylim = c(-90,90), pch = 16,
asp = 1,
main = "plot(r, useRaster=TRUE) -> \nincorrect y-axis limits")
useRaster=FALSE
## Chain of functions dispatched by `plot(r, useRaster=FALSE)`
getMethod("plot", c("RasterLayer", "missing"))
raster:::.plotraster2
raster:::.imageplot
image.default
## Call within image.default() that sets up the plotting region
plot(NA, NA, xlim = xlim, ylim = ylim, type = "n", xaxs = xaxs,
yaxs = yaxs, xlab = xlab, ylab = ylab, ...)
## Example showing that the above call produces the wrong aspect ratio
plot(c(-180,180), c(-90,90),
xlim = c(-180,180), ylim = c(-90,90), pch = 16,
main = "plot(r,useRaster=FALSE) -> \nincorrect aspect ratio")
Man, I got stumped and ended up just turning the foreground color off to plot. Then you can take advantage of the fact that the raster plot method calls fields:::image.plot, which lets you just plot the legend (a second time, this time showing the ink!). This is inelegant, but worked in this case:
par("fg" = NA)
plot(r, xlim = c(xmin(r), xmax(r)), ylim = c(ymin(r), ymax(r)), axes = FALSE)
par(new = TRUE,"fg" = "black")
plot(r, xlim = c(xmin(r), xmax(r)), ylim = c(ymin(r), ymax(r)), axes = FALSE, legend.only = TRUE)
axis(1, pos = -90, xpd = TRUE)
rect(-180,-90,180,90,xpd = TRUE)
ticks <- (ymin(r):ymax(r))[(ymin(r):ymax(r)) %% 20 == 0]
segments(xmin(r),ticks,xmin(r)-5,ticks, xpd = TRUE)
text(xmin(r),ticks,ticks,xpd=TRUE,pos=2)
title("sorry, this could probably be done in some more elegant way")
This is way I solved this problem
require(raster)
r = raster()
# default for raster is 180 row by 360 cols = 64800 cells
# fill with some values to make more interesting
r[]= runif(64800, 1, 1000)
# Set margin for text
par(mar=c(2, 6, 6, 2))
# Set some controls for the raster cell colours and legend
MyBrks<-c(0,1,4,16,64,256,1E20)
MyLbls<-c("<1","<4","<16","<64","<256","<Max")
MyClrs<-c("blue","cyan","yellow","pink","purple","red")
# Plot raster without axes or box or legend
# Note xlim and ylim don't seem do much unless you want to trim x and y
plot(r,
col=MyClrs,
axes=FALSE,
box=FALSE,
legend=FALSE
)
# Set up the ranges and intervals for axes - you can get the min max
# using xmin(r) and ymax(r) and so on if you like
MyXFrm <- -180
MyXTo <- 180
MyXStp <- 60
MyYFrm <- -90
MyYTo <- 90
MyYStp <- 30
# Plot the axes
axis(1,tick=TRUE,pos=ymin(r),las=1,at=seq(MyXFrm,MyXTo ,MyXStp ))
axis(2,tick=TRUE,pos=xmin(r),las=1,at=seq(MyYFrm ,MyYTo ,MyYStp ))
# Plot the legend use xpd to plot the legend outside the plot region
par(xpd=TRUE)
legend(MyXTo ,MyYTo ,
legend=MyLbls[1:6],
col= MyClrs,
fill=Clrs[1:6],
bg=rgb(0,0,0,0.85),
cex=0.9,
text.col="white",
text.font=2,
border=NA
)
# Add some axis labels and a title
text(-220,0,"Y",font=2)
text(0,-130,"X",font=2)
text(0,120,"My Raster",font=4,cex=1.5)
I think the best (or simplest) solution is to use image():
library(raster)
# Raster
r = raster()
r[]= rnorm(ncell(r))
# Spatial points
x = c(-100, 0, 100)
y = c(100, 0, 100)
points = SpatialPoints(data.frame(x,y))
# plot
image(r)
plot(points, add=T, pch=16, cex=2)