Intro to R

• About R
• R Basics
• Working With Data
• Hands-On Project
• Further Learning

• for newbies.
  • If you already have some experience in R, please raise your hands.
• about R's programming aspects.
  • It's designed to help you start using and coding R.
• not about Statistics.
  • I assume that you already have some basic knowledge of Statistics
• for you to start learning R afterwards.
  • This workshop is designed to give you a nudge to learn more R down the road. There are many resources to learn R in depth.

By the end of this this workshop, you will be able to:

• Install R libraries/packages.
• Import and export data from/to a simple CSV file in R.
• Distinguish the differences between R data objects and convert them.
• Use subsetting methods to create subsamples.
• Create basic plots using both base plot in R.
• Create R markdown documents and R slides.
• Create tidy data and manipulate those data using dplyr package.

• R is a statical programming language/environment.
• R is open source and free.
• R is widely used.
• R is cross-platform.
• R is hard to learn (really?).

• S: R's ancestor
• S-Plus: Commercial; modern implementation of S
• SAS: Commercial; widely used in the commercial analytics.
• STATA: Commercial; widely used by economists.
• SPSS: Commercial; easy to use; widely used in Social Science.
• MATLAB: Commercial; can do some Stats.
• Python: Also can do some Stats; one of the two mostly used programming languages in data science.

• R community is active and constantly growing
• R is one of the most popular stat programming lang
• R has tons of user generated libraries/packages
• R code is easily shared with others
• R is constantly improved

• R community is active and constantly growing
• R is one of the most popular stat tools/programming languages
• R has tons of user generated libraries/packages
• R code is easily shared with others
• R is constantly improved

Further reading about R's popularity in science and engineering:

R moves up to 5th place in IEEE language rankings at https://www.r-bloggers.com/r-moves-up-to-5th-place-in-ieee-language-rankings/

https://spectrum.ieee.org/static/interactive-the-top-programming-languages-2017 (6th in 2017)

IEEE Spectrum: Popular Programming Languages in 2017

• R community is active and constantly growing
• R is one of the most popular stat tools/programming languages
• R has tons of user generated libraries/packages
• R code is easily shared with others
• R is constantly improved

On the Internet:

• Google! (Now Google knows “r” means the R programming language)
• Stack Overflow: http://stackoverflow.com/questions/tagged/r
• Cross-Validated: the statistics Q&A site http://stats.stackexchange.com/
• Rseek meta search engine: http://rseek.org/
• R-help listserv: https://www.r-project.org/mail.html

Let's get ready for R-ing!

Open your RStudio app. Open your browser.

• Upon opening RStudio, you will see the Console
  • 

• You can perform some calculations by typing into it.

2+2

[1] 4

• You can also use variables and assign them values.

x = 2
x+2

[1] 4

# The "<-" operator is the same as "="
y <- x+1 # "<-" is preferred
x+y

[1] 5

• When you create a variable it will show up on the environment pane in RStudio

• Operators are symbols that use some input and operate on them.
• Arithmetic operators are our common mathematical symbols.

x*x # Multiplication

[1] 4

x/y # Division

[1] 0.6666667

x^3 # Exponentiation

[1] 8

• Logical operators output either True or False

2 < 5  # Less Than

[1] TRUE

x >= 2 # Greater Than Or Equal To

[1] TRUE

x == 1 # Equal to. NOTE: This is different from "="

[1] FALSE

x != x # Not Equal to

[1] FALSE

• R can work with more than just numbers.

class("hello, world!") # The class function returns the data type of a variable or value.

[1] "character"

class(x == y) 

[1] "logical"

• Up until now, we have been working with single numbers and variables also known as scalars.
• In R, there are also vectors which are sequences of values.

a <- c(1,2,3)
b <- c(3:1)

for(i in 1:3) {
  print(a[i]+b[i])
}

[1] 4
[1] 4
[1] 4

• R's basic data object is “vector”
• Operators are optimized for vector (element-wise) operations
• Use vectorized operations than loops whenever possible

a + b

[1] 4 4 4

a^2

[1] 1 4 9

• Functions take an input and produce an output.
• R contains many functions that are built-in, some of which we have already seen such as class() and c().

abs(-5) # Absolute Value

[1] 5

sqrt(4) # Square Root

[1] 2

sum(a)  # Sum

[1] 6

toupper("hello") # To Upper Case

[1] "HELLO"

sd(a) # Can you guess what this funtion calc?

[1] 1

cor(a,b) # How about this?

[1] -1

• R is known for its community and its libraries/packages
• Packages are collections of R functions, data, and compiled code in a well-defined format. The directory where packages are stored is called the library.
• We will install the dplyr package, a popular package for data manipulation.
• install.packages('dplyr')

library('dplyr')

• With all these packages and functions you may forget how to use one or need to remember which one to use.

# This will search for usage of the hist function
?hist

# This will search packages and functions about histograms
??histogram

• Review of some basic R features
• Open ./exercises/intro/exercise_1.Rmd (this is a R markdown file)

• A data frame is used for storing data tables. It is a list of vectors of equal length.

names <- c("John", "Kim", "Terry")
ages  <- c(21, 34, 16)
male  <- c(TRUE, FALSE, TRUE)
data.frame(names, ages, male)

  names ages  male
1  John   21  TRUE
2   Kim   34 FALSE
3 Terry   16  TRUE

• R's datasets package has some built in datasets that we will be using.
• The CO2 data frame has 84 rows and 5 columns of data from an experiment on the cold tolerance of the grass species Echinochloa crus-galli.

help(CO2)

• You can use '$' to access a vector (column/feature/etc) of a dataframe

# The head function shows the first few entries in a dataframe.
head(CO2)

  Plant   Type  Treatment conc uptake
1   Qn1 Quebec nonchilled   95   16.0
2   Qn1 Quebec nonchilled  175   30.4
3   Qn1 Quebec nonchilled  250   34.8
4   Qn1 Quebec nonchilled  350   37.2
5   Qn1 Quebec nonchilled  500   35.3
6   Qn1 Quebec nonchilled  675   39.2

# The summary function shows summary statistics.
summary(CO2)

     Plant             Type         Treatment       conc     
 Qn1    : 7   Quebec     :42   nonchilled:42   Min.   :  95  
 Qn2    : 7   Mississippi:42   chilled   :42   1st Qu.: 175  
 Qn3    : 7                                    Median : 350  
 Qc1    : 7                                    Mean   : 435  
 Qc3    : 7                                    3rd Qu.: 675  
 Qc2    : 7                                    Max.   :1000  
 (Other):42                                                  
     uptake     
 Min.   : 7.70  
 1st Qu.:17.90  
 Median :28.30  
 Mean   :27.21  
 3rd Qu.:37.12  
 Max.   :45.50  

hist(CO2$uptake) # Use help function for more plotting options

boxplot(CO2$conc)

plot(CO2$uptake, CO2$conc)

• Review of datasets and data visualization
• Open ./exercises/intro/exercise_2.Rmd

Operators that can be used to extract subsets of R objects.

• '[' and ']' always returns an object of the same class as the original; can be used to select more than one element.
• '[[' and ']]' is used to extract elements of a list or a data frame; it can only be used to extract a single element.
• $ is used to extract elements of a list or data frame by name.

x <- c("a", "b", "c", "c", "d", "a")
x[1]
x[1:4]
x[x > "a"] 
u <- x > "a" # what's u here?
u
x[u] # subsetting using a boolean vector
y <- list(foo=x, bar=x[u]) 
y
y[[1]]
y$bar

• Create a dataframe containing the first 10 observations.
• Create a dataframe where only the Plant identifier is Qn3.
• Create a dataframe only containing observations with the CO2 concentration level greater than 600.

co2_10obs <- CO2[<row>, <col>]
co2_Qn3 <- CO2[?, ?]
co2_gt600conc <- subset(CO2, conc > 600)

• Dplyr is the most common package used for data exploration and transformation
• Dplyr functions: filter, select, arrange, mutate, summarise (plus group_by)

head(filter(CO2, Treatment=='nonchilled'))

  Plant   Type  Treatment conc uptake
1   Qn1 Quebec nonchilled   95   16.0
2   Qn1 Quebec nonchilled  175   30.4
3   Qn1 Quebec nonchilled  250   34.8
4   Qn1 Quebec nonchilled  350   37.2
5   Qn1 Quebec nonchilled  500   35.3
6   Qn1 Quebec nonchilled  675   39.2

• select will output only the columns that you choose

head(select(CO2, Plant, conc, uptake), 3)

  Plant conc uptake
1   Qn1   95   16.0
2   Qn1  175   30.4
3   Qn1  250   34.8

• You can chain/pipe dplyr functions together
  • The infix (or pipe) operator '%>%' will feed the resulting object into the 1st paramater of the next function

x <- select(filter(CO2, Treatment=='nonchilled'), Plant, conc, uptake) # OR
y <- CO2 %>% filter(Treatment=='nonchilled') %>% select(Plant, conc, uptake)
x == y

      Plant conc uptake
 [1,]  TRUE TRUE   TRUE
 [2,]  TRUE TRUE   TRUE
 [3,]  TRUE TRUE   TRUE
 [4,]  TRUE TRUE   TRUE
 [5,]  TRUE TRUE   TRUE
 [6,]  TRUE TRUE   TRUE
 [7,]  TRUE TRUE   TRUE
 [8,]  TRUE TRUE   TRUE
 [9,]  TRUE TRUE   TRUE
[10,]  TRUE TRUE   TRUE
[11,]  TRUE TRUE   TRUE
[12,]  TRUE TRUE   TRUE
[13,]  TRUE TRUE   TRUE
[14,]  TRUE TRUE   TRUE
[15,]  TRUE TRUE   TRUE
[16,]  TRUE TRUE   TRUE
[17,]  TRUE TRUE   TRUE
[18,]  TRUE TRUE   TRUE
[19,]  TRUE TRUE   TRUE
[20,]  TRUE TRUE   TRUE
[21,]  TRUE TRUE   TRUE
[22,]  TRUE TRUE   TRUE
[23,]  TRUE TRUE   TRUE
[24,]  TRUE TRUE   TRUE
[25,]  TRUE TRUE   TRUE
[26,]  TRUE TRUE   TRUE
[27,]  TRUE TRUE   TRUE
[28,]  TRUE TRUE   TRUE
[29,]  TRUE TRUE   TRUE
[30,]  TRUE TRUE   TRUE
[31,]  TRUE TRUE   TRUE
[32,]  TRUE TRUE   TRUE
[33,]  TRUE TRUE   TRUE
[34,]  TRUE TRUE   TRUE
[35,]  TRUE TRUE   TRUE
[36,]  TRUE TRUE   TRUE
[37,]  TRUE TRUE   TRUE
[38,]  TRUE TRUE   TRUE
[39,]  TRUE TRUE   TRUE
[40,]  TRUE TRUE   TRUE
[41,]  TRUE TRUE   TRUE
[42,]  TRUE TRUE   TRUE

• arrange() sorts rows by a column

# Use desc() to sort descending
CO2 %>% arrange(desc(uptake)) %>% head(3)

  Plant   Type  Treatment conc uptake
1   Qn3 Quebec nonchilled 1000   45.5
2   Qn2 Quebec nonchilled 1000   44.3
3   Qn3 Quebec nonchilled  675   43.9

• mutate() creates new variables

CO2 %>% mutate(conc_L = conc / 1000) %>% head()

  Plant   Type  Treatment conc uptake conc_L
1   Qn1 Quebec nonchilled   95   16.0  0.095
2   Qn1 Quebec nonchilled  175   30.4  0.175
3   Qn1 Quebec nonchilled  250   34.8  0.250
4   Qn1 Quebec nonchilled  350   37.2  0.350
5   Qn1 Quebec nonchilled  500   35.3  0.500
6   Qn1 Quebec nonchilled  675   39.2  0.675

• group_by() and summarise() work together to aggregate variables

CO2 %>% group_by(Type) %>% summarise(avg_uptake = mean(uptake))

# A tibble: 2 x 2
         Type avg_uptake
       <fctr>      <dbl>
1      Quebec   33.54286
2 Mississippi   20.88333

• In datasets with a lot of columns you can select all columns but selected ones by using the '-' operator select(CO2, -Plant)
• When aggregating data you need an aggregate function such as mean, median, n (number of rows in a group), sum, etc.
• If your data has missing values, you need to add the parameter na.rm = TRUE, this will skip any missing values.
• When working with data, the data manipulation process often takes more time than the analysis.

• Review of some data manipulation functions in dplyr
• Open ./exercises/intro/exercise_3.Rmd

• Beginner/Intermediate Topics
  • User Defined Functions
  • Control Structures
  • Statistical Methods
  • Data Cleaning
  • Data Visualization Libraries (ggplot2, plotly)
    • https://plot.ly/r/dashboard/

• Advanced Topics
  • Deep Learning
  • Functional Programming
  • Object Oriented Programming
  • Debugging Tools
  • Performance

• In this workshop we have been using the console but when working on a project you will generally use the editor so you can save your scripts. This helps anyone reproduce your findings.
• When using a script, you must run your code either all at once (Run All) or Line-by-Line

• Download the data to your computer at http://bit.ly/r-hands-on-s2018
• You can import the data to R either by code or by clicking Import Dataset on the Environment Tab

# You need to put the full path to the file in quotes or change the working directory [getwd(), setwd()]
Births <- read.csv("Births2015.csv")
Auto   <- read.csv("Auto.csv")

1. What is the name of the 1st car in the dataset?
2. What is the total number of babies born in 2015?
3. Make a boxplot of car mpg.
4. Make a histogram of number of births.
5. How many babies are born on Wednesdays?
6. Which date had the least amount of babies born?
7. Is there a relationship between the number of cylinders and mpg?
8. Plot the average mpg for cars of each number of cylinders?

1. head(Auto,1): chevrolet chevelle malibu
2. sum(Births$births): 3978497
3. boxplot(Auto$mpg)
4. hist(Births$births)
5. Births %>% group_by(wday) %>% summarise(sum=sum(births)) %>% arrange(sum): 638513
6. arrange(Births, births): 12/25/2015
7. cor(Auto$mpg, Auto$cylinders) or plot(Auto$cylinders, Auto$mpg)

• Swirl (http://swirlstats.com/)
• DataCamp (https://www.datacamp.com/courses/tech:r)
• R for Data Science by Hadley Wickham (http://r4ds.had.co.nz/) [FREE]
• Statistics.com (https://www.statistics.com/landing-page/r-courses/)