SPATIAL ANALYSIS WITH R

The R spatial analysis packages include spatial point processing, spatial autocorrelation, smoothing, interpolation, geostatistics, etc. The “sp” package in R offers a wide variety of geostatistical functions such as foundation classes, interface to coordinate systems, utility plotting methods, sampling methods, etc. Other R geospatial packages are listed in Table II.

Table II: R Spatial Analysis Programs

1. R as a Calculator: R language use the usual arithmetic operators and the data type used is modes. The modes are logical, numeric and complex numbers. One of the simple possible tasks in R is to enter an arithmetic operation and receive a result. For example, if we want to add two numbers then we can type in the terminal
   >2+2
   >4

2. Assigning value to variables: R has symbolic variables like any other programming language that is used to represent value to the assigned variable. The operator “<-” is known as assignment operator and it assigns the value of the expression on right to the object on left. For example, we can assign 5 to the variable x which can be used for subsequent arithmetic expression.
   >x<-5
   and then type
   >x+x
   to get 10 as our final output

3. Methods of data entry: R can handle entire data vectors as single objects and there are various ways of inputting data in R. We can directly type the values in the command line with concatenation function “c”. Data can be entered one at a time from the keyboard using scan or “read.table” command.
   w<- c(60, 72, 57, 90, 95, 72)
   y <- scan( )
         >data<-read.table(“read_my_file.txt”,header=T)
   If the data is separated by “tab” or “space” then it can be specified in the command
   >data<-read.table(“read_my_file.txt”,sep=“/t”,header=T)
   >attach (data)
   >names(data)

4. Arithmetic operations: In R many complicated calculations can be done by using addition, substraction, multiplication, division and exponentiation as  operator.
   >x<-3+8
   >5^2-5*2
   [1] 15 will appear as the outcome. Its better to specify the order of evaluation of expression by using parenthesis and no space is required to separate components in an arithmetic operation.
   >1-3*3
   [1] -8

5. Useful Builtin Function: Many calculations are already built into the program [2] and results can be easily obtained by using these functions as listed in Table III.
   Table III: R built in functions.

   Sl. No.	Function	Purpose
   1	mean	to find average
   2	sd	to find standard deviation
   3	max	to obtain maximum element in a data set
   4	min	to find minimum element in a data set
   5	range	to find range of values
   6	median	to find median of the data set
   7	var	to find variance of the data set
   8	sort	command for sorting in increasing order of magnitude
   9	diff	creates a vector of differences by substracting  previous  element from next  element.
   10	cov(a,b)	covariance between a and b
   11	cor(a,b)	correlation between a and b
   12	length(a)	length of a vector
   13	floor(x)	round down to next lowest integer
   14	ceil(x)	round up next largest integer
   15	round(x)	round to nearest integer
   16	anova	analysis of variance
   17	summary (x)	summarises the data (x)
   18	assign(“x”,3)	assigns the value 3 to object x
   19	c()	combines or concatenates terms together
   20	rep(x, each=4)	repeats x 4 times

6. Graphics with R: R provides flexible and powerful graphical analysis. There are a number of commands which helps in obtaining simple graphs through R by using the command “plot” and giving various specifications about the appearance of the graph in both x and y axis. Table IV lists some of the standard plot functions.

A simple example is to plot a sine curve. First the intervals are defined along with the number of data points to support the curve.
>x <- seq(-2*pi, 2*pi, len = 100)
>x
>str(x)
>summary(x)
Then we can plot it (the type parameter specifies the line   type):
matplot(x, sin(x), type=“l”)
To get an idea about the various “matplot()” options, run:
>?matplot
You can see some “matplot()” examples by running:
>example(matplot)
More examples follows:
>barplot(table(x),xlab=“DigitalNumbers”,ylab=“Frequency”,col“gray70”); where x is the data which is already being inputed in R
> plot(c,xlab=“BANDS”,ylab=“Digital Number”)

> boxplot(a,xlab=“BANDS”,ylab=“Digital    Number) where a is the data set which R statistical software has already read.


Figure 1: Line graph generated in R.


Figure 2: Quadratic line graph obtained in R


Figure 3: Line with quadratic, cubic, power -1 and power -2 generated in R



Figure 4: Bar plot


Figure 5: Boxplot

Table IV: Standardr plot functions

Sl. No.	Function	Name of the plot
1	plot()	Scatterplot
2	hist()	Histogram
3	boxplot()	Box-and-whiskers plot
4	stripchart()	Stripchart
5	barplot()	Bar diagram
6	stem()	Stem-and-leaf display

7. Saving, Storing and Retrieving Work in R: When we quit R session, we should tupe “yes” to “save workspace image”. When we restart R all the data and variables from the previous session are available.

8. Getting help in R: To get help in R use “help.search” function with the query in double quotes “ ”,  for example help.search “data input ”.

9. Add-on (CRAN) Packages in R: R has numerous addon packages for different applications. Packages can be downloaded for the CRAN Mirror site (http://cran.r-project.org/) and installed in R. Table V prensents some of the R packages. More packages can be found on http://cran.r-project.org/. If we want to install packages then the simple command is “install.packages” or from the Linux Terminal type: $R CMD INSTALL package_name.tar.gz. For example, “R CMD INSTALL spgrass6 _0.3-7.tar.gz”.

   Table V: Standard packages in R.

   Sl. No.	Package	Purpose
   1	nortest	Normality test
   2	randomforest	Classifier
   3	cluster	Clustering tool
   4	rgdal	Geospatial abstraction data library in R
   5	raster	Geographic analysis and modelling with raster data
   6	kohonen	Supervised and unsupervised self-organising maps
   7	tseries	Time-series data analysis