| Basic things about R | Reading data in | Basic things about geoR |
To quit R type at the prompt:
> q()
And the following message will appear:
Save workspace image? [y/n/c]:
Type y to save your session, n will NOT save the session and c will send you back to the R session.
Let's now start using R.
Restart R from your desktop menu.
R saves your data in a file called .Rdata in the current working directory.
File --> Change Dir
and type in the directory where you want R to save your data during this session.
Note: If you wish you can create different desktop icons for different R projects. To do so create an R desktop shortcut for each project and change the option Start in in the shortcut properties.
Hands on: Now let's see how R handles data using the idea of objects.
Examine the examples below.
n1 <- 15
n1
messa <- "Hello you"
messa
n2 <- 1:4
n2
country <- c("France", "Spain", "Italy")
country
ls()
ls(pattern="n")
ls.str()
rm(n1)
ls()
rm(list=ls())
ls()
args(ls)
args(anova)
For a more detailed explanation of the function arguments and capabilities use the function help:
help(ls)
help(anova)
You can choose to visualize the help in your browser (html help). To do so, before using the help() type:
options(htmlhelp = TRUE) # windows only
help.start()
After typing this inspect the page shown in your browser and its links.
Ask for help on specific functions as before and the documentation will now be shown in your browser:
help(ls)
help(anova)
Vectors
x1 <- 10
x1
x2 <- c(1, 3, 6)
x2
x2[1]
x2[2]
length(x2)
is.vector(x2)
is.matrix(x2)
is.numeric(x2)
is.character(x2)
x3 <- 1:10
x3
x4 <- seq(0,1, by=0.1)
x4
x4[x4 > 0.5]
x4 > 0.5
x5 <- seq(0,1, l=11)
x5
x6 <- rep(1, 5)
x6
x7 <- rep(c(1, 2), c(3, 5))
x7
x8 <- rep(1:3, rep(5,3))
x8
x9 <- rnorm(10, mean=70, sd=10)
x9
sum(x9)
mean(x9)
var(x9)
min(x9)
max(x9)
summary(1:10)
x10 <- x9[x9 > 72]
Inspect and try the arguments of the functons seq and rep:
args(seq) args(rep)Matrices and arrays
m1 <- matrix(1:12, ncol=3)
m1
length(m1)
dim(m1)
nrow(m1)
ncol(m1)
m1[1,2]
m1[2,2]
m1[,2]
m1[3,]
dimnames(m1)
dimnames(m1) <- list(c("L1", "L2", "L3","L4"), c("C1","C2","C3"))
dimnames(m1)
m1[c("L1","L3"),]
m1[c(1,3),]
m2 <- cbind(1:5, 6:10)
m2
m3 <- cbind(1:5, 6)
m3
ar1 <- array(1:24, dim=c(3,4,2))
ar1[,2:3,]
ar1[2,,1]
sum(ar1[,,1])
sum(ar1[1:2,,1])
Data-Frames
d1 <- data.frame(X = 1:10, Y = c(51, 54, 61, 67, 68, 75, 77, 75, 80, 82))
d1
names(d1)
d1$X
d1$Y
plot(d1)
plot(d1$X, d1$Y)
d2 <- data.frame(Y= c(10+rnorm(5, sd=2), 16+rnorm(5, sd=2), 14+rnorm(5, sd=2)))
d2$lev <- gl(3,5)
d2
by(d2$Y, d2$lev, summary)
d3 <- expand.grid(1:3, 4:5)
Lists
l1 <- list(A=1:10, B="THIS IS A MESSAGE", C=matrix(1:9, ncol=3))
l1
l2 <- lm(Y ~ X, data=d1)
l2
is.list(l2)
class(l2)
summary(l2)
anova(l2)
names(l2)
l2$pred
l2$res
plot(l2)
l3 <- aov(Y ~ lev, data=d2)
l3
summary(l3)
Other useful objects: ts, factor
An very special type of object: functions
Several examples already mentioned above: lm, plot, plot.default, min, max, sum, etc.
lm
plot
plot.default
but this is not always immediatly available, for example:
min
max
lines
A simple way to read data in is by using ASCII (text) files).
Example:
Cruciani.dat Cruciani.border
cru <- read.table("Cruciani.dat")
cru
Note:In Windows you can also use the menu to do this.
For what follows you will need the packagegeoR installed in your computer.
The package geoR is an official contributed package, which means that it is available
at CRAN but it is not included in the basic distribution of R.
The package is also available for download from the geoR web-page.
To download and install/upgrade the latest version follow the intructions in the web page:
http://www.leg.ufpr.br/~paulojus/geoR/geoRdoc/installpackage.html
After the install, load the package geoR by typing:
library(geoR)or, if installed in a non-default location,
library(geoR, lib="PATH_TO_YOUR_geoR_DIRECTORY")
The package has a function to read data in called read.geodata,
which is a modification of the standard read.table function
to prepare objects of the "class" geodata.
Therefore we read the data again as well as the file with the borders of the study area:
cru <- read.geodata("Cruciani.dat", head=T, coords.col=2:3, data.col=4)
cru.borda <- read.table("Cruciani.border", head=T)[,2:3]
cru.borda <- rbind(cru.borda, cru.borda[1,])
cru
is.list(cru)
class(cru)
plot(cru, bord=cru.borda)
Notice that, because cru is an object of the class geodata the function being executed in the last command is plot.geodata.
args(plot.geodata)
help.start()
help(plot.geodata)
Try to use/change the arguments for the plot function. For example:
plot(cru, bord=cru.borda, lambda=0)
Another descriptive function available is points.geodata:
points(cru, bord=cru.borda)
args(points.geodata)
points(cru, bord=cru.borda, cex.min=1, cex.max=1, pt.div="quartile")
points(cru, bord=cru.borda, cex.min=1, cex.max=1,col=gray(seq(0.9,0,l=length(cru$data))))
points(cru, bord=cru.borda, cex.min=1, cex.max=1, col=gray(seq(0.9,0,l=length(cru$data))), xla="Coord X", ylab="Coord Y")
points(cru, lambda=0, bord=cru.borda)
points(cru, lambda=0, bord=cru.borda, cex.min=1, cex.max=1, pt.div="quartile")
points(cru, lambda=0, bord=cru.borda, cex.min=1, cex.max=1, col=gray(seq(0.9,0,l=length(cru$data))))
hist(cru$data, main="", xlab="K")
hist(log(cru$data), main="", xlab="log(K)")
There are a few data-set included in the package. Examples:
data(wolfcamp)
wolfcamp
help(wolfcamp)
data(s100)
data(parana)
EXERCISE: Explore these data set using the functions above and/or others.