Programming$in$R - About people.tamu.edupeople.tamu.edu/~alawing/materials/ESSM689/Programming.pdfESSM689$Quan,tave$Methods$in$ Ecology,$Evolu,on$and$Biogeography$ Ecosystem$Science$and$Management|$

ESSM 689 Quan,ta,ve Methods in Ecology, Evolu,on and Biogeography

Ecosystem Science and Management | Texas A&M University (c) 2015, A. Michelle Lawing

Programming in R

A. Michelle Lawing Ecosystem Science and Management Texas A&M University College Sta,on, TX 77843 [email protected]



Schedule for today •  My presenta,on

–  Review –  New stuff

•  Model Selec,on and Bootstrapping presenta,on

•  Model Selec,on and Bootstrapping demonstra,on

•  BREAK

•  Model Selec,on and Bootstrapping tutorial



Review: Probability and Distribu,on

•  R makes it easy to draw probability distribu,ons to demonstrate sta,s,cal concepts

•  Func,ons for each distribu,on follow a par,cular format

Name Descrip,on

dname() Density or probability func,on

pname() Cumula,ve density func,on

qname() Quan,le func,on

rname() Random deviates



Review: Probability and Distribu,on

Normal Uniform Student t dnorm(0) dunif(0) dt(0,20) pnorm(0) punif(0) pt(0,20) qnorm(0) qunif(0) qt(0,20) rnorm(1) runif(1) rt(1,20)



−4 −2 0 2 4

0.0

0.4

0.8

Comparison of Distributions

x value

Den

sity

− P

roba

bilit

y Distributionsdf=1df=3df=8df=30normunif



−4 −2 0 2 4

0.0

0.1

0.2

0.3

0.4

Comparison of t Distributions

x value

Den

sity

− P

roba

bilit

y Distributionsdf=1df=3df=8df=30norm



Normal Distribution

x values

Den

sity

− P

roba

bilit

y

P( −0.674 < x values < 0.674 ) = 0.5

−4 −2 0 2 4

0.0

0.1

0.2

0.3

0.4



Student T Distribution Two−Tailed

x values

Den

sity

− P

roba

bilit

y

P( −1.44 > x values < 1.44 ) = 0.15

−4 −2 0 2 4

0.0

0.1

0.2

0.3

0.4



Programming in R



Ge`ng Help

•  There are various control func,ons that are used to establish the control flow

help(Control)





Execu,ng a file

•  A set of commands can be saved in a file and executed from the command line in R

source("hep://michellelawing.info/file.R") #source a file, url, or other connec,on help(source)



Execu,ng a file

source("hep://michellelawing.info/file.R", echo = T) source("hep://michellelawing.info/file.R", print.eval = T) source("hep://michellelawing.info/file.R", print.eval = F) source("hep://michellelawing.info/file.R", verbose = T) TIP: Open R from right clicking on your script file



Where is the file?

source(“notThere.R”) #some,mes R can’t find the file #R searches the current director

getwd() #returns the working directory dir() #lists the files in the wd setwd() #change wd



if statement

x <-‐ 1 #assign x a number if( x < 5) { #if x is less than 5

x <-‐ x + 1 #then increment by 1 }



else statement


x <-‐ x + 1 #then increment by 1 } else { #in all other cases

x <-‐ x – 1 #decrement by 1 }



chain ifs together


x <-‐ x + 1 #then increment by 1 } if else (x == 5){ #or if x equals 5

x <-‐ x + 1 #then increment by 1 } else { #in all other cases

x <-‐ x – 1 #decrement by 1 }



for statements for(i in 1:10) { #loop the temp var i 10 ,mes

print(i) #print i each loop } for(i in seq(0, 1, by = 0.1)) { #loop the i from 0 to 1 by 0.1

print(i) #print i each loop } for(i in 10:5) { #loop i from 10 to 5

print(i) #print i each loop }



Random Walk

e <-‐ 200 #assign the number of ,mes steps y <-‐ 0 #assign the star,ng value for(i in 2:e) { #loop i from 1 to e y[i] <-‐ rnorm(1, mean = y[i – 1])

#draw a random number } plot(1:e, y, pch = 16, type = "l”) TIP: You don’t have to ini,alize arrays in R.



while statement

x <-‐ 1 #assign a star,ng value cntr <-‐ 0 #set the counter to 0 while(x < 3) { #repeat un,l condi,al is met x <-‐ rnorm(1) #draw random number cntr <-‐ cntr + 1 #increment the counter } cntr #print the number of draws



Other useful statements

•  repeat –  Same as while, except evaluates at least once –  Need to define stopping point

•  break –  break stops the execu,on of the current loop

•  next –  next skips the remaining statements within a loop and starts the next looping cycle



func,ons

foo <-‐ func,on ( a, b) { #assign a func,on name x <-‐ runif( 10, a, b) #draw 10 values from a unif dist mean(x) #calculate the mean

} foo( -‐1, 1) #execute func,on with min and max input

#the last expression in the func,on is returned



func,ons – return

foo <-‐ func,on ( a, b) { x <-‐ runif( 10, a, b) return(list(mu = mean(x), std = sd(x)))

} #use return statement foo( -‐1, 1) #returns list



func,ons – return mul,ple

foo <-‐ func,on ( a, b) { x <-‐ runif( 10, a, b) return(list(a = a, b = b, mu = mean(x), std = sd(x)))

} #use return statement foo( -‐1, 1) #returns list and input



func,ons – return plot

foo <-‐ func,on ( a, b) { x <-‐ runif( 10, a, b) hist(x) return(list(a = a, b = b, mu = mean(x), std = sd(x)))

} #use return statement foo( -‐1, 1) #returns list and input



Exercise -‐ Programming

Create a func,on that plots a random walk by using rnorm. The func,on should read in the ini,al trait value and the number of ,me steps.

Programming$in$R - About people.tamu.edupeople.tamu.edu/~alawing/materials/ESSM689/Programming.pdfESSM689$Quan,tave$Methods$in$ Ecology,$Evolu,on$and$Biogeography$ Ecosystem$Science$and$Management|$

Documents