gstlearn

Introduction

gstlearn is a package developed by the Geostatistical Team of Mines ParisTech. It contains several well established geostatistical functions and methods and some prototypes based on ongoing research.

Loading gstlearn

In each R session, you must load the gstlearn library

R used as a pocket calculator

1+3

## [1] 4

exp(log(sqrt(2^2)))

## [1] 2

Assignments

a = 1
b <- 3
a + b

## [1] 4

d = a + b
d

## [1] 4

d = 7
d

## [1] 7

Working directory

ls()
rm(d)
ls()

Quit

q()

Save the workspace (y).

The content of your session is saved in the file .Rdata (with the .Rhistory) of your working directory.

Caution: these files are hidden.

To continue, you must relaunch R (and load gstlearn again).

Classes in R

b <- 3
class(b)

## [1] "numeric"

col = "blue"
class(col)

## [1] "character"

reponse=FALSE
class(reponse)

## [1] "logical"

Booleans

a = 1
a < 2

## [1] TRUE

comp = a < 2
comp

## [1] TRUE

comp & reponse

## [1] FALSE

comp | reponse

## [1] TRUE

!comp

## [1] FALSE

a = 1
a == 1

## [1] TRUE

a != 1

## [1] FALSE

a > 1

## [1] FALSE

a >= 1

## [1] TRUE

Vectors

x = c(1,9,2,9,4,5)
x

## [1] 1 9 2 9 4 5

class(x)

## [1] "numeric"

x[1]

## [1] 1

x[c(1,3)]

## [1] 1 2

x[-2]

## [1] 1 2 9 4 5

2:5*3

## [1]  6  9 12 15

x = c(1,9,2,9,4,5)
x[2:5]

## [1] 9 2 9 4

selection = x > 2
selection

## [1] FALSE  TRUE FALSE  TRUE  TRUE  TRUE

x[selection]

## [1] 9 9 4 5

Length of a vector

length(x)

## [1] 6

Matrices

M = matrix(1:20,nrow=4,ncol=5)
M[2,1]

## [1] 2

M[2,1]=3
M[2,1]

## [1] 3

M[,1]

## [1] 1 3 3 4

M[2,]

## [1]  3  6 10 14 18

dim(M)

## [1] 4 5

Data frame

data = data.frame(M)
class(data)

## [1] "data.frame"

data[,3]

## [1]  9 10 11 12

names(data)

## [1] "X1" "X2" "X3" "X4" "X5"

names(data) = c("A","B","C","D","E")
data

##   A B  C  D  E
## 1 1 5  9 13 17
## 2 3 6 10 14 18
## 3 3 7 11 15 19
## 4 4 8 12 16 20

data

##   A B  C  D  E
## 1 1 5  9 13 17
## 2 3 6 10 14 18
## 3 3 7 11 15 19
## 4 4 8 12 16 20

data$C

## [1]  9 10 11 12

data[data$A>1,]

##   A B  C  D  E
## 2 3 6 10 14 18
## 3 3 7 11 15 19
## 4 4 8 12 16 20

Data Frame from a File

Download the ASCII file called Scotland_Temperatures.csv and store it on your disk in the current working directory.

filename = "Scotland_Temperatures.csv"
if(flagInternetAvailable){
  download.file(paste0("https://soft.minesparis.psl.eu/gstlearn/data/Scotland/",filename), filename, quiet=TRUE)
}

temperatures = read.csv(filename,header=TRUE,na="MISS")
class(temperatures)
names(temperatures)

Some functions

set.seed(123)
rnorm(3)

## [1] -0.5604756 -0.2301775  1.5587083

x=c(3,4,7,4,2)
x

## [1] 3 4 7 4 2

sort(x)

## [1] 2 3 4 4 7

unique(x)

## [1] 3 4 7 2

x=c(3,4,7,4,2)
sum(x)

## [1] 20

mean(x)

## [1] 4

var(x)

## [1] 3.5

sd(x)

## [1] 1.870829

Help of Functions

Search for help on the function image.

?image

To get some help on a class :

class?matrix

Some functions of interest

Function outer

A = c(2,4,10)
B = c(3,7)
outer(A,B,"-")

##      [,1] [,2]
## [1,]   -1   -5
## [2,]    1   -3
## [3,]    7    3

outer(A,B,"*")

##      [,1] [,2]
## [1,]    6   14
## [2,]   12   28
## [3,]   30   70

Loops

Dimension and initialize a vector: numeric

Loop from 1 to 10: store loop index in a vector; then print the vector

tab = numeric(10)
for (i in 1:10)
{
  tab[i] = i
}
print(tab)

##  [1]  1  2  3  4  5  6  7  8  9 10

Create your own function

Generate a linear transform of the input argument

my.func = function(x,a=2,b=4)
{
  y = a * x + b
  y
}

Play my function

x = seq(0,20)
y = my.func(x)
plot(x,y,main="Plotting my function")