Master [R]

PoisotLab/data-trek-2020master
This repo is mounted by: R

1. Data types, data structures and indexing

1.1 Basics

Object, assignment, functions, how to comment and get help

x
0.4s
R
R
x <- 2
0.2s
R
R
x + 2
0.5s
R
R
x * 3 # we can use R as a calculator!
0.3s
R
R
log(1) # Functions help us execute things; we usually have to provide arguments
0.2s
R
R
?log # Don't know how a function works? Ask for help!
0.6s
R
R

1.2 Data types

Numeric (integers or doubles)

my_numeric <- c(x, x+2 , x *3) # Naming objects tips
0.1s
R
R

Character

my_character <- c("blue", "bleu", "azul", "hyacinthum")
0.1s
R
R

Boolean

is_french <- c(FALSE, TRUE, FALSE, FALSE)
0.2s
R
R

1.3 Data classes

Vectors

Ordered collection of elements.

my_character
0.4s
R
R

What happens when we add 1 to a logical vector?

is_french + 1
0.4s
R
R

Matrix

dim(my_vector) <- c(lines, columns)			# reshapes my_vector
my_character2 <- my_character
dim(my_character2) <- c(2,2)
0.2s
R
R
my_character2
0.1s
R
R
0 items

Lists

list(name_element1 = element1, name_element2 = element2, name_element3 = element3)
my_list <- list(word = my_character, french = is_french)
my_list
0.5s
R
R

Data frames

data.frame(vector1, vector2)	# bind vectors with the same length
my_df <- data.frame(word = my_character, french = is_french)
my_df
0.2s
R
R
0 items

1.4 Dimensions

dim(matrix)
length(object)

Can you guess the value of the length of my_df?

dim(my_df)
length(my_df)
dim(my_list)
length(my_list)
0.4s
R
R

1.5 Indexing

vector[index]
matrix[row, column]
list[[element]]
my_character[2]
0.4s
R
R
my_df[2,1]
0.4s
R
R
my_list[[1]]
0.4s
R
R
my_list[[1]][2]
0.4s
R
R

1.6 Slicing

my_matrix <- matrix(c(34, 9, 6, 5, 3, 50, 43, 27, 98, 100), nrow=5)
0.2s
R
R
my_matrix[my_matrix[,1]>5,]
0.2s
R
R
0 items

2. Files

Absolute paths

C:/Users/RonBumblefootThal/Documents/RFolder/MyFirstProject/Draft/IDon'tKnowWhatI'mDoing/etc.R

Relative Paths

~/I_love_my_project/CoolCode.R

2.1 Working directories

dir()
0.5s
R
R
setwd("data-trek-2020")
0.2s
R
R
getwd()
0.7s
R
R
list.files()
0.4s
R
R

2.2 Save/write files

Data frame example

soa_tour <- data.frame(country = c("USA", "UK", "FRA", "GER", "BRA"), 
                       frequency = c(34, 9, 6, 5, 3), 
                       continents = c("north_america", "europe", "europe", 
                                      "europe", "south_america"))
0.3s
R
R
write.csv(object, path)
write.csv(soa_tour, file="data/clean/soa_tour.csv")
0.2s
R
R
list.files(data/clean)
Shift+Enter to run
R
R

2.3 Load/read files

From your PC

object <- read.csv(path)
soa_tour <- read.csv(file="data/clean/soa_tour.csv")
0.2s
R
R

From url

object <- read.csv(url("http://remote.repo/data/file.csv"))

Metabolic rates data: http://sciencecomputing.io/data/metabolicrates.csv

metabolic_rates <- read.csv(url("http://sciencecomputing.io/data/metabolicrates.csv"))
0.3s
R
R

From url to your PC, then read

download.file(url, destfile)
download.file(url = "http://sciencecomputing.io/data/metabolicrates.csv",
              destfile = "data/raw/metabolicrates.csv")
0.5s
R
R
metabolic_rates <- read.csv("data/raw/metabolicrates.csv")
0.2s
R
R

3. Control Flow

You already apply control flow when you decide how to go to work during winter.

For example:

  • You take the metro if it's snowy

  • You take the metro if it's cold

  • You walk every other time

Now, let's put that into code!

3.1 Conditional evaluation

Simple if statement

Structure:

if (condition is true) {
	do expression
}

Example:

weather <- "snowy"
0.3s
R
R
if (weather == "snowy") {
  print("Take the metro!")
}
0.3s
R
R
weather <- "clear"
0.1s
R
R
if (weather == "snowy") {
  print("Take the metro!")
}
0.2s
R
R

If/else statement

Structure:

if (condition) {
	expression 1
 } else { 
 	expression 2
}

Example:

if (weather == "snowy") {
  print("Take the metro!")
} else {
  print("Let's walk!")
}
0.5s
R
R

Nested if /else statement

if (condition 1) {
	expression 1
  if (condition 2) {
    expression 2
  }
}

Example:

temperature <- -15
0.2s
R
R
if (weather == "snowy") {
  print("Take the metro!")
} else {
  if (temperature < -20) {
  	print("Take the metro!")
	} else {
    print("Let's walk!")
  }
}
0.6s
R
R

Best practice:

if (weather == "snowy") {
  print("Take the metro!")
} else if (temperature < -20) {
  print("Take the metro!")
} else {
  print("Let's walk!")
}
0.3s
R
R

Adding a condition: 

if (weather == "snowy" | temperature < -20) {
  print("Take the metro!")
} else {
  print("Let's walk")
}
0.4s
R
R

3.2 For loops

Simple for loops

Using for loops, you can then plan your schedule for a few days.

What we had:

weather <- "snowy"
temperature <- -15
0.2s
R
R

But what about this?

weather_vec <- c("snowy", "cloudy", "snowy", "clear", "rainy")
temperature_vec <- c(-15, -23, -2, -40, 5)
0.2s
R
R

Does the same code work?

if (weather_vec == "snowy" | temperature_vec < -20) {
  print("Take the metro!")
} else {
  print("Let's walk")
}
0.4s
R
R

Iterations

for (i in iterations) {
	content of the for loop
}
for(i in 1:5){
	 print(temperature_vec[i] + 2)
}
0.3s
R
R

More generally:

length(temperature_vec)
0.4s
R
R
for(i in 1:length(temperature_vec)){
	print(temperature_vec[i] + 2)
}
0.4s
R
R

If statement inside for loops

Structure:

for (i in iterations) {
	if (condition) {
 	 	expression1
  } else {
  	expression2
  }
}

Example: 

# Previous statement
if(weather_vec == "snowy" | temperature_vec < -20){
  	print("Take the metro!")
} else {
 	 	print("Just walk")
}
Shift+Enter to run
R
R

for (i in 1:length(weather_vec)){
	# Previous statement
	if(weather_vec == "snowy" | temperature_vec < -20){
  		print("Take the metro!")
	} else {
 	 		print("Just walk")
	}
}
0.3s
R
R
# Will this work?
for (i in 1:length(weather_vec)){
  if(weather_vec == "snowy" | temperature_vec < -20){
  	print("Take the metro!")
	} else {
 	 	print("Just walk")
	}
}
0.3s
R
R

3.3 Extras

Some logical operators

Comparisons:

  • less than (<)

  • more than (>)

  • less than or equal to (<=)

  • more than or equal to (>=)

  • equal to (==)

  • not equal to (!=)

Logic:

  • not x (! x)

  • x or y (x | y)

  • x and y (x & y)

4. Functions

4.1 Syntax and arguments

Basic syntax

functionname <- function(argument1, argument2) { # Name and arguments 
	result = expression		# What the function does
  return(result) # What the function returns
}	
temp_difference <- function(temp1, temp2) { # Name and arguments
  result = temp2 - temp1 # What the function does
  return(result) # What the function returns
}
0.2s
R
R

# Apply on values
temp_difference(-5, -15)
0.5s
R
R

# Apply on variables
temperature <- c(-15, -23)
temp_difference(temperature[1], temperature[2])
0.5s
R
R

Calling (personal) functions within functions

absolute_temp_difference <- function(var1, var2) {
  result <- temp_difference(var1, var2)
  abs_result <- abs(result)
  return(abs_result)
}
0.2s
R
R
absolute_temp_difference(temperature[1], temperature[2])
0.3s
R
R

4.2 Scope

Variables can exist either in global or local scope.

Remember, the element to return in our function was called `abs_result`. 

# What will this return, outside the function?
result
0.2s
R
R

Here is a second example for ecologists who like to count living things:

# global variables
trees <- 4 
squirrels <- 10
0.1s
R
R

count_living_things <- function() {
  birds <- 5 # local variable
  squirrels <- 20
  return(c(birds, squirrels, trees))
}
count_living_things() # global and local variables returned
0.4s
R
R

birds # does not exist in global scope
0.2s
R
R

squirrels
0.4s
R
R

4.3 Integration

Combining functions and control flow

Let's come back to our previous example about transportation according to the weather.

Here is the forecast for the week and the weekend:

# Week forecast
weather_week <- c("snowy", "cloudy", "snowy", "clear", "rainy")
temperature_week <- c(-15.0, -23.0, -2.0, -40.0, 5.0)
# Weekend forecast
weather_weekend <- c("snowy", "rainy")
temperature_weekend <- c(-3.0, 2.0)
0.1s
R
R

Now, let's build a function that will work with either the week or weekend forecasts.

It will look like:

transportation <- function {
	for (all days of the week/weekend) {
		if (snowy or cold) {
 				 take metro
    } else {
    		 walk
    }
 }
for (i in 1:length(weather_vec)) {
 	if (weather_vec[i] == "snowy" || temperature_vec[i] < -15.0) {
    print("Take the metro")
  } else {
    print("Just walk")
  }   
}
0.4s
R
R
choose_transportation <- function(weather, temperature) {
  for (i in 1:length(weather)) {
    if (weather[i] == "snowy" || temperature[i] < -15.0) {
      print("Take the metreo!")
    } else {
      print("Just walk")
    }
   }
 }
0.2s
R
R
# Plan for the week
choose_transportation(weather_week, temperature_week)
0.3s
R
R
# Plan for the weekend
choose_transportation(weather_weekend, temperature_weekend)
0.3s
R
R

4.4 Exercise - Planning the week

Exercise to integrate the following:

  • Functions

  • Control flow

  • Files

  1. Write a function to read a file if it exists, downloading it first if it does not exist.

  2. Apply the choose_transportation function to the data in the file

# Pseudocode
function (file, url)
	if (file exists)
  	read file
  else
  	download file
    read file
library("R.utils", quietly = TRUE)
0.2s
R
R
# Useful functions
?file.exists     # library "R.Utils"
?read.csv
?download.file
Forecast data url: http://bit.ly/dt-forecast
read_if_exists <- function(filename, url) {
  if (file.exists(filename)) {
    read.csv(filename)
  } else {
    download.file(url, filename)
    read_if_exists(filename, url)
  }
}
0.2s
R
R
filename = "forecast.csv"
url = "http://bit.ly/dt-forecast"
read_if_exists(filename, url)
0.2s
R
R
0 items
choose_transportation(forecast$weather, forecast$temperature)
Shift+Enter to run
R
R

Extras

Default values

# Define function
add_and_multiply <- function(var1, var2, var3 = 1) {
  result <- (var1 + var2) * var3
  return(result)
}
0.2s
R
R
multiply(1.0, 2.0, 2.0)
0.4s
R
R

Runtimes (1)