![BASICS OF R LANGUAGE
R is case sensitive.
Counting starts at 1 (x[1] is the first element of
x).
class(x) will return the class (type) of x.](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-31-320.jpg)
![BASICS OF R LANGUAGE - VECTOR
R’s basic unit is the vector (column).
It is one dimensional, with a length.
To create a vector: x <- c(1,2,3,4) [the arrow
is the ‘assign’ statement]
A single element (e.g., created by ‘X <- 1’) is a
vector of length 1.
All elements of a vector have to be of the
same type (numeric, logical, Boolean).](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-32-320.jpg)
![BASICS OF R LANGUAGE - LIST
A list is an ordered set of elements, which can
be of different types.
An element can be itself a list.
To create a list:
y <-list(1,2,3,'four', c(6,7))
class(y) returns [1] "list“
For many purposes, a list is a single element,
unless something iterates through it.](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-33-320.jpg)
![ACCESSING ELEMENTS OF A VECTOR
(HTTPS://WWW.R-BLOGGERS.COM/R-ACCESSORS-EXPLAINED/
WILL EXPLAIN WHAT SORT OF OBJECTS ARE RETURNED)
For a vector, single brackets can be used:
class(x) returns: [1] "numeric“
x[2] returns: [1] 2 the [1] means ‘first element’.
Elements can also be accessed by Boolean logic:
x <- c(1,2,3,4)
x >3 returns: [1] FALSE FALSE FALSE TRUE
x[x>3] returns: [1] 4
class(x) [1] "numeric"
[1] 2](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-35-320.jpg)
![ACCESSING ELEMENTS OF A LIST
Since each element of a list can itself be a list, it’s a bit
more complicated.
Using single brackets on a list returns a list.
Using double brackets on a list returns a vector.
y[1] returns a list, with the first element being 1.
y[[1]] returns a vector, with the first element being 1.](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-36-320.jpg)
![ACCESSING ELEMENTS OF A DATA
FRAME
The classic way to select a column (vector, variable,
field) of a data frame is using the ‘$’ operator:
data_xz$x will return x, which is the column ‘1,2,3,4’.
You can also use brackets for [row, column]:
data_xz[1,2] returns the first row, second column,
data_xz[1,] returns the first row, all columns (and vice
versa for row).
data_xz[data_xz$x>2,] would return the rows of data_xz
where x > 2.](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-37-320.jpg)
![EDA
To find out what the column types of a data
frame/tibble are, use the command: lapply(sheet1,
class) .
‘Lapply’ is member of a set of ‘apply’ functions with
different prefixes, which iterate a function through an
object, and then return one of variety of object types
back]
To see the first/last few rows, use ‘head()’ and ‘tail()’.
Because sheet1 is a tibble, those functions will return a
tibble (a subset of the original tibble).](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-45-320.jpg)
![GGPLOT EXAMPLE
For this example, we will use the mpg data frame,
which comes with the ggplot2 package (most
packages have sample data frames).
It is directly accessed by ‘ggplot2::mpg
(package::item, which also works for functions within
packages).
First, issue the ggplot() command, specifying the data
set: ggplot(data=mpg) + [the ‘+’ means continuing]
Second, specify a GEOM (what to plot) and a AES
(aesthetic – what columns to use)](https://image.slidesharecdn.com/introductiontorstudioonaws20200506-200509153627/85/Introduction-to-r-studio-on-aws-2020-05_06-50-320.jpg)
Introduction to r studio on aws 2020 05_06
- 1. INTRODUCTION TO RSTUDIO ON AWS Barry DeCicco Ann Arbor Tech Meetup, May 6, 2020
- 2. CONTENTS Setting up the AWS Instance. Orientation on RStudio.
- 4. RSTUDIO ON AWS You need to use the right AMI on AWS, one set up for RStudio. Set that up, and connect to Dropbox.
- 7. SELECT ‘NEXT’ ON EACH STEP AND ACCEPT THE DEFAULTS UNTIL YOU HIT ‘CONFIGURE SECURITY GROUP’
- 22. USING RSTUDIO
- 23. This is where you can write and save commands . Console – issue commands; output appears here. Environment – list of objects, including data sets. Multip-purpose windows: List of packages, File structure, Help, Plot Viewer Menu Bar
- 24. When ‘Files’ is selected, you can navigate through the file system.
- 25. When ‘Plots’ is selected, you can see the plots which you have created. (it’s empty now, but will have plots later).
- 26. When ‘Packages’ is selected, you can see your installed packages (think python modules) and see information and whether or not they are loaded.
- 27. When ‘Help’ is selected, you access help resources.
- 28. PACKAGES R comes with a base set of packages. You can think of packages as similar to modules in python. To obtain packages, you must INSTALL them. They usually install from a CRAN mirror (CRAN is the Comprehensive R Archive Network, the hosting system for R). To use packages, you must LOAD them, using the library() command.
- 29. PACKAGES (CON.) To list installed packages, type: installed.packages() To load the list into a data set for ease of use, type: package_list <- as.data.frame(installed.packages()) installed.packages()
- 30. You can browse data frames here. Objects are listed here. You can type commands into the console.
- 31. BASICS OF R LANGUAGE R is case sensitive. Counting starts at 1 (x[1] is the first element of x). class(x) will return the class (type) of x.
- 32. BASICS OF R LANGUAGE - VECTOR R’s basic unit is the vector (column). It is one dimensional, with a length. To create a vector: x <- c(1,2,3,4) [the arrow is the ‘assign’ statement] A single element (e.g., created by ‘X <- 1’) is a vector of length 1. All elements of a vector have to be of the same type (numeric, logical, Boolean).
- 33. BASICS OF R LANGUAGE - LIST A list is an ordered set of elements, which can be of different types. An element can be itself a list. To create a list: y <-list(1,2,3,'four', c(6,7)) class(y) returns [1] "list“ For many purposes, a list is a single element, unless something iterates through it.
- 34. BASICS OF R LANGUAGE – DATA FRAME A data frame is a list of vectors, with the same length. The vectors can be of different types. To create a data frame: Import a file. Coerce another type of object (i.e., a matrix). Combine some vectors: z <- c(1,4,9,16) data_xz <- data.frame(x,z)
- 35. ACCESSING ELEMENTS OF A VECTOR (HTTPS://WWW.R-BLOGGERS.COM/R-ACCESSORS-EXPLAINED/ WILL EXPLAIN WHAT SORT OF OBJECTS ARE RETURNED) For a vector, single brackets can be used: class(x) returns: [1] "numeric“ x[2] returns: [1] 2 the [1] means ‘first element’. Elements can also be accessed by Boolean logic: x <- c(1,2,3,4) x >3 returns: [1] FALSE FALSE FALSE TRUE x[x>3] returns: [1] 4 class(x) [1] "numeric" [1] 2
- 36. ACCESSING ELEMENTS OF A LIST Since each element of a list can itself be a list, it’s a bit more complicated. Using single brackets on a list returns a list. Using double brackets on a list returns a vector. y[1] returns a list, with the first element being 1. y[[1]] returns a vector, with the first element being 1.
- 37. ACCESSING ELEMENTS OF A DATA FRAME The classic way to select a column (vector, variable, field) of a data frame is using the ‘$’ operator: data_xz$x will return x, which is the column ‘1,2,3,4’. You can also use brackets for [row, column]: data_xz[1,2] returns the first row, second column, data_xz[1,] returns the first row, all columns (and vice versa for row). data_xz[data_xz$x>2,] would return the rows of data_xz where x > 2.
- 38. IMPORTING DATA There is a package ‘tidyverse’ which contains a number of useful data manipulation and plotting commands; it is the heart of common modern R usage. To load it, give the command ‘library(tidyverse’). You will get some notes as it loads, including what commands from other packages are masked. Masked means that a command ‘doit’ from a package A might be overridden by another command with the same name, ‘doit’, from package B.
- 39. IMPORTING DATA (CON) The command ‘read_csv’ is used to import data from .csv files. It has many optional arguments; in this case they are not needed. To create a data frame called ‘sheet1’: sheet1 <- read_csv("/home/rstudio/Dropbox/Barry DeCicco's shared workspace/Documents/R Studio Projects/7 Visualizations You Should Learn in R/Big Mart Datasets/Big Mart Dataset - Sheet1.csv“)
- 40. Command echoed here, along with specifications chosen by r Command run from here Data frame ‘sheet1’ listed in objects
- 41. POKING AND PRODDING OBJECTS In the upper right-hand window (‘Global Environment’) , objects are listed, in categories such as ‘Data’ and ‘Values (variables, lists, etc.) . To view them, you can click on the object name for data, or use the command: View(object). A spreadsheet-like view will pop-up. You can filter and sort this view.
- 42. POKING AND PRODDING OBJECTS
- 43. POKING AND PRODDING OBJECTS (CON) To find out what type of object an object is, use the command: class(object).: Class(x) will return ‘numeric’, since x is a numeric vector. Dim(sheet1) will return the number of rows and columns. Length(y) will return the length of the list y. Length(sheet1) will return the number of columns of sheet1 (since a data frame is a list of vectors, which are the coumns). Typing an object’s name will cause R to print that object out, in varying degrees of usefulness. The output will depend on the class of the object.
- 44. Since the tidyverse command ‘read_csv()’ was used to import the .csv file, ‘sheet1’ is a Tibble, which is a data frame with extra features. One of those is better printing:
- 45. EDA To find out what the column types of a data frame/tibble are, use the command: lapply(sheet1, class) . ‘Lapply’ is member of a set of ‘apply’ functions with different prefixes, which iterate a function through an object, and then return one of variety of object types back] To see the first/last few rows, use ‘head()’ and ‘tail()’. Because sheet1 is a tibble, those functions will return a tibble (a subset of the original tibble).
- 46. EDA str(x) will return the ‘structure’ of x, which is a set of parts and names for those parts. To see the first/last few rows, use ‘head()’ and ‘tail()’. Because sheet1 is a tibble, those functions will return a tibble (a subset of the original tibble).
- 47. Summary(object) will return a summary, which will vary depending on the class of the object.
- 48. str(object) will return the ‘structure’ of the object, which is a set of parts and names for those parts.
- 49. GRAPHS ➢ In the Long Long Ago Time, the reason that people started using R was for its graphics. ➢ The most common graphics package is ggplot2, which is based on the book ‘The Grammar of Graphs’. ➢ The method is to build a graph up of basic commands and modifiers. This can allow exploration. ➢ These graph objects can be stored, then have commands add to them, and then be printed. ➢ There are many, many, many extension packages to ggplot2.
- 50. GGPLOT EXAMPLE For this example, we will use the mpg data frame, which comes with the ggplot2 package (most packages have sample data frames). It is directly accessed by ‘ggplot2::mpg (package::item, which also works for functions within packages). First, issue the ggplot() command, specifying the data set: ggplot(data=mpg) + [the ‘+’ means continuing] Second, specify a GEOM (what to plot) and a AES (aesthetic – what columns to use)
- 51. ggplot(data = mpg) + + geom_point(mapping = aes(x = displ, y = hwy))
- 52. These plots can have more features, either from the start or added on later: ggplot(data = mpg) + geom_point(mapping = aes(x = displ, y = hwy, color = class))
- 53. ggplot(data = mpg) + geom_point(mapping = aes(x = displ, y = hwy, color = class)) + facet_wrap(~ cyl, nrow = 2)
- 54. ONE ADDITIONAL THING Most output from functions can be assigned to objects: > model_for_mpg = lm(cty ~ cyl +trans, data=ggplot2::mpg) > model_for_mpg Call: lm(formula = cty ~ cyl + trans, data = ggplot2::mpg) Coefficients: (Intercept) cyl transauto(l3) transauto(l4) transauto(l5) transauto(l6) 30.469 -2.013 -1.416 -2.159 -2.536 -2.039 transauto(s4) transauto(s5) transauto(s6) transmanual(m5) transmanual(m6) -1.065 -1.056 -1.014 -1.144 -1.495
- 55. ONE ADDITIONAL THING ➢ At this point there is an object containing the linear model results. ➢ This has parts, which can be extracted as variables or objects. ➢ There are packages (such as ‘broom’ which can convert this into a data frame, allowing the analysis of sets of models, and each presentations of results,
- 56. RMARKDOWN AND KNITR ➢ Everything until now has been done in an r script file. ➢ You can also create an Rmarkdown file (File > New File > R Markdown). ➢ These files mix code chunks, text and the output from the code chunks. ➢ The resulting output can be in .rtf, .html or .pdf (extra setup is required for the latter, on your desktop). ➢ The code chunks can be in over 20 languages!
- 57. ‘Knit’ button causes document to be generated Block of text Code chunk – can be printed or not in the final document.
- 59. QUESTIONS?
- 60. REFERENCES ‘Advanced Data Analysis from an Elementary Point of View’ (a statistics class taught in R): https://www.stat.cmu.edu/~cshalizi/ADAfaEPoV/ R for Data Science https://r4ds.had.co.nz/ The tidyverse: https://www.tidyverse.org/ Examples of ggplot graphs: Top 50 ggplot2 Visualizations - The Master List (With Full R Code): http://r-statistics.co/Top50-Ggplot2-Visualizations-MasterList-R- Code.html Dynamic Documents with R and Knitr (by Yihui Xie) https://yihui.name/knitr/
- 61. REFERENCES (CON.) Getting R (on your own computer): https://cran.r-project.org/ Getting Rstudio (on your own computer): https://rstudio.com/