03 the r language part 2

7/24/2019 03 the R Language Part 2

1/163

2014 RStudio, Inc. All rights reserved. Follow @rstudioapp

All Training materials are provided "as is" and withoutwarranty and RStudio disclaims any and all express and

implied warranties including without limitation theimplied warranties of title, fitness for a particularpurpose, merchantability and noninfringement.

The Training Materials are licensed under the CreativeCommons Attribution-Noncommercial 3.0 United StatesLicense. To view a copy of this license, visithttp://creativecommons.org/licenses/by-nc/3.0/us/or send aletter to Creative Commons, 171 Second Street, Suite300, San Francisco, California, 94105, USA.
http://creativecommons.org/licenses/by-nc/3.0/us/http://twitter.com/rstudioapp


2/163

Studio

2014 RStudio, Inc. Follow @rstudioapp

Garrett GrolemundMaster Instructor, RStudio

August 2014

Retrieve and use informationin precise, e#cient ways

The R language (2 of 2)
http://twitter.com/rstudioapp


3/163

2014 RStudio, In

1. Subsetting

2. R Packages

3. Logical tests

4. Missing values

Studio


4/163

2014 RStudio, In

How can you save just the fifth element of xto y

How can you change the fifth element of x to a

Question

x


5/163

Subsetting


6/163 2014 RStudio, In

With your neighbor, run the code on thefollowing slide IN YOUR HEADS

Your turn

vec



vec df

6 1 3 6 10 5John

Paul

George

Ringo

1940

1940

1942

1943

guitar

guitar

drums

bass

name birth instrument

vec[2]

vec[c(5, 6)]

vec[-c(5,6)]

vec[vec > 5]

# Predict what the following code will do

# DON'T RUN IT!

df[c(2, 4), 3]

df[ , 1]

df[ , "instrument"]

df$instrument



Subset notation

Studio

vec[2]

name of objectto subset



Subset notation

Studio

vec[?]


brackets(brackets always mean

subset)


10/163 2014 RStudio, In

Subset notation

Studio

vec[?]


brackets(brackets always mean

subset)

an index(that tells R which

elements to include)

d


11/163

2014 RStudio, In

Each dimension needs its own index!

Studio

vec[?] 6 1 3 6 10 5

S di


12/163

2014 RStudio, In


Studio

vec[?] 6 1 3 6 10 5

S di


13/163

2014 RStudio, In


Studio

vec[?]

df[?,?]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

St di


14/163

2014 RStudio, In


Studio

vec[?]

df[?,?]

whichrowstoinclude

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

St dio


15/163

2014 RStudio, In


Studio

vec[?]

df[?,?]

whichrowstoinclude

whichcolumnsto include

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

Studio


16/163

2014 RStudio, In


Studio

vec[?]

separatedimensions

with acomma

df[?,?]

whichrowstoinclude

whichcolumnsto include

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

Studio


17/163

2014 RStudio, In


Studio

vec[?]

df[?,?]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

But what should go in the indexes?

Studio


18/163

2014 RStudio, In

Four ways to subset

1. Integers

2. Blank spaces

3. Names

4. Logical vectors (TRUE and FALSE)

Studio

Studio


19/163

2014 RStudio, In

Integers (positive)

Positive integers behave just like ijnotation inlinear algebra

Studio

df[?,?]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

Studio


20/163

2014 RStudio, In

Integers (positive)

Studio

df[2,?]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Studio


21/163

2014 RStudio, In

Integers (positive)

Studio

df[2,3]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Studio


22/163

2014 RStudio, In

Integers (positive)

df[2,3]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Studio


23/163

2014 RStudio, In

df[c(2?4),c(2?3)]

Integers (positive)

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Studio


24/163

2014 RStudio, In

df[c(2,4),c(2?3)]

Integers (positive)

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Studio


25/163

2014 RStudio, In

df[c(2,4),c(2,3)]

Integers (positive)

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Studio


26/163

2014 RStudio, In

Integers (positive)

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[c(2,4),c(2,3)]


Studio


27/163

2014 RStudio, In

Integers (positive)


John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[c(2,4),3]

?

Studio


28/163

2014 RStudio, In

Integers (positive)


John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[c(2,4),3]

Studio


29/163

2014 RStudio, In

1. Colons are a useful way to create vectors

1:4

# 1 2 3 4

df[1:4, 1:2]

2. Repeating input repeats outputdf[c(1,1,1,2,2), 1:3]

Studio


30/163

2014 RStudio, In

Integers (zero)

As an index, zero will return nothingfrom adimension. This creates an empty object.

vec[0]

# numeric(0)

df[1:2, 0]

# data frame with 0 columns and 2 rows

Studio


31/163

2014 RStudio, In

Integers (negative)

Negative integers return everything but theelements at the specified locations.

You cannot use both negative and positiveintegers in the samedimension

Studio


32/163

2014 RStudio, In

Integers (negative)



vec[c(5,6)] 6 1 3 6 10 5

Studio


33/163

2014 RStudio, In

vec[-c(5,6)]

Integers (negative)



6 1 3 6 10 5

Studio


34/163

2014 RStudio, In

Integers (negative)



John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[c(2:4), 2:3]

Studio


35/163

2014 RStudio, In

Integers (negative)



John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[-c(2:4), 2:3]

Studio


36/163

2014 RStudio, In

Integers (negative)



John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[-c(2:4),-(2:3)]


37/163

2014 RStudio, In

Your Turn

1. Fix these poorly written subset commands

vec(1:4)

vec[-1:4]vec[3, 4, 5]

Studio


38/163

2014 RStudio, In

() for functions, [] for subsetting

vec[1:4]

# 6 1 3 6

Dont mix positive and negative integers; distribute thnegative sign (e.g., -1:4 = -1 0 1 2 3 4).

vec[-(1:4)]

# 10 5

Pass multiple values for the same dimension as a ve

vec[c(3, 4, 5)]

# 3 6 10


39/163

2014 RStudio, In

Your Turn

What is wrong with these subsetting commandsWhat will they do?

mat[2]

df[1]


40/163

2014 RStudio, In

1

2

3

4

5

6

7

8

9

mat[2]

How R makes a matrix


41/163

2014 RStudio, In


1 2 3 4 5 6 7 8 9vec



42/163

2014 RStudio, In

How R makes a matrix1

2

3

4

5

6

7

8

9

vec



43/163

2014 RStudio, In


1

23

vec

4

5

6

7

8

9



44/163

2014 RStudio, In


1

23

4

56

7

8

9

vec



45/163

2014 RStudio, In


1

23

4

56

7

89

vec



46/163

2014 RStudio, In


1

23

4

56

7

89

vecmatrix


47/163

2014 RStudio, In

1

2

3

4

5

6

7

8

9

1 2 3 4 5 6 7 8 9

mat[2]


48/163

2014 RStudio, In

John

df[2]

Paul

George

Ringo

1940

1942

1943

1940

bass

drums

guitar

guitar

How R makes a data frame


49/163

2014 RStudio, In

List c("a","b","c","d") c(1, 2, 3, 4) c(T, F, T, F)

How R makes a data frame


50/163

2014 RStudio, In

List c("a",

"b",

"c",

"d")

c(

1,

2,

3,

4)

c(T,

F,

T,

F)


51/163

2014 RStudio, In

c(

"a",

"b",

"c",

"d")

c(

1,

2,

3,

4)

c(T,

F,

T,

F)

List


52/163

2014 RStudio, In

c(

"a",

"b",

"c",

"d")

c(

1,

2,

3,

4)

c(T,

F,

T,

F)

Listdata frame


53/163

2014 RStudio, In

John

df[2]

Paul

George

Ringo

1940

1942

1943

1940

bass

drums

guitar

guitar

c("guitar", "bass",

"guitar","drums")

c("John","Paul",

"George","Ringo")

c(1940, 1942,

1943, 1940)

Studio


54/163

2014 RStudio, In

Blank spaces

Blank spaces return everything

(i.e., no subsetting occurs on that dimension)

vec[ ] 6 1 3 6 510

Studio


55/163

2014 RStudio, In

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

Blank spaces



df[1,]

Studio


56/163

2014 RStudio, In

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

Blank spaces



df[ ,2]

Studio


57/163

2014 RStudio, In

If your object has names, you can ask forelements or columns back by name.

Names

vec[c("a","b","d")] 6 1 3 6 510

Studio


58/163

2014 RStudio, In


names(vec)


59/163

2014 RStudio, In


names(vec)


60/163

2014 RStudio, In


names(vec)


61/163

2014 RStudio, In

Names


John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[ ,"birth"]


N

Studio


62/163

2014 RStudio, In

Names


John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

df[ ,c("name","birth")]


L i l

Studio


63/163

2014 RStudio, In

You can subset with a logical vector of the samelength as the dimension you are subsetting.Each element that corresponds to a TRUE willbe returned.

Logical

vec[c(FALSE,TRUE,FALSE,TRUE,TRUE,FALSE)]

6 1 3 6 510

L i l

Studio


64/163

2014 RStudio, In


Logical

vec[c(FALSE,TRUE,FALSE,TRUE,TRUE,FALSE)]

6 1 3 6 510

c(FALSE,TRUE,FALSE,TRUE,TRUE,FALSE)

L i l

Studio


65/163

2014 RStudio, In

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


Logical

df[c(FALSE,TRUE,TRUE,FALSE), ]

L i l

Studio


66/163

2014 RStudio, In

You can subset with a logical vector of the same

length as the dimension you are subsetting.Each element that corresponds to a TRUE willbe returned.

Logical

df[c(FALSE,TRUE,TRUE,FALSE), ]

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass

c(FALSE,TRUE,TRUE,FALSE)

Studio

Subset notation


67/163

2014 RStudio, In

Subset notation

Your Turn


68/163

2014 RStudio, In

Your Turn

Write down as many ways to extract the name"John"from dfas you can. Make sure eachworks. You have two minutes.

John

Paul

George

Ringo

1940

1940

1941

1943

guitar

guitar

drums

bass


# Answers

Studio


69/163

2014 RStudio, In

df[1, 1]

df[1, "name"]

df[1, -(2:3)]

df[1, c(TRUE, FALSE, FALSE)]

df[-(2:4), 1]

df[-(2:4), "name"]

df[-(2:4), -(2:3)]

df[-(2:4), c(TRUE, FALSE, FALSE)]df[c(TRUE, FALSE, FALSE, FALSE), 1]

df[c(TRUE, FALSE, FALSE, FALSE), "name"]

df[c(TRUE, FALSE, FALSE, FALSE), -(2:3)]

df[c(T, F, F, F), c(T, F, F)]

Your Turn


70/163

2014 RStudio, In

Your Turn

lst


71/163

2014 RStudio, In

Subsetting lists

sum(lst[1]) # Error!

# What is the di"

erence?

lst[c(1,2)]

lst[1]

lst[[1]]

lst c(1, 2) c("a", "b", "c")TRUE

Studio


72/163

2014 RStudio, In

If list xis a train carryingobjects, then x[[5]]is the

object in car 5; x[4:6]is a

train of cars 4-6.

http://twitter.com/#!/RLangTip/status/118339256388304896

Studio
http://twitter.com/#!/RLangTip/status/118339256388304896


73/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "cTRUE

Studio


74/163

2014 RStudio, In

c(1, 2) c("a", "b", "cTRUE

Studio


75/163

2014 RStudio, In

c("a", "b", "cc(1, 2) TRUE

lst[c(1,2)]

Studio


76/163

2014 RStudio, In

lst[c(1,2)]

c(1, 2) c("a", "b", "cTRUE

c(1, 2) TRUE

# [[1]]

# [1] 1 2

#

# [[2]]

# [1] TRUE

Studio


77/163

2014 RStudio, In

c("a", "b", "cc(1, 2) TRUE

lst[1]

Studio


78/163

2014 RStudio, In

c("a", "b", "cTRUEc(1, 2)

c(1, 2)lst[1]

# [[1]]

# [1] 1 2

Studio


79/163

2014 RStudio, In

lst[[1]]

c("a", "b", "cc(1, 2) TRUE

Studio


80/163

2014 RStudio, In

c("a", "b", "cc(1, 2) TRUE

c(1, 2)lst[[1]]

Studio


81/163

2014 RStudio, In

c("a", "b", "cc(1, 2) TRUE

Wha

twillt h

lst[[1]][2]

Studio


82/163

2014 RStudio, In

lst[[1]][2]

c("a", "b", "cc(1, 2) TRUE

Studio


83/163

2014 RStudio, In

lst[[1]][2]

c("a", "b", "cc(1, 2) TRUE

c(1, 2)[2]

Studio

(" " "b" "( )


84/163

2014 RStudio, In

lst[[1]][2]

c("a", "b", "cc(1, 2) TRUE

c(1, 2)[2]

2

$

Studio


85/163

2014 RStudio, In

$The most common syntax for subsetting listsand data frames

names(lst)


86/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "c")TRUE

lst$alpha

Studio

alpha beta gamma


87/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "c")TRUE

lst$alpha

name of list

Studio

alpha beta gamma


88/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "c")TRUE

lst$alpha

$name of list

Studio

alpha beta gamma


89/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "c")TRUE

lst$alpha

$name of listname of element

(no quotes)

Studio

alpha beta gamma


90/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "c")TRUE

lst$alpha

$name of listname of element

(no quotes)

c(1, 2)

Studio

alpha beta gamma


91/163

2014 RStudio, In

lst c(1, 2) c("a", "b", "c")TRUE

lst$alpha

c(1, 2)

Same

aslst

Studio

John 1940 guitar



92/163

2014 RStudio, In

df$birth

Paul

George

Ringo 1940

1941

1943 guitar

drums

bass

Studio

John 1940 guitar



93/163

2014 RStudio, In

df$birth

name of dataframe

Paul

George

Ringo 1940

1941

1943 guitar

drums

bass

Studio

John 1940 guitar



94/163

2014 RStudio, In

df$birth

$name of data

frame

Paul

George

Ringo 1940

1941

1943 guitar

drums

bass

Studio

John 1940 guitar



95/163

2014 RStudio, In

df$birth

$name of data

framename of column

(no quotes)

Paul

George

Ringo 1940

1941

1943 guitar

drums

bass

Studio

John 1940 guitar



96/163

2014 RStudio, In

df$birth

$name of data

framename of column

(no quotes)

c(1940, 1941, 1943, 1940)

Paul

George

Ringo 1940

1941

1943 guitar

drums

bass


97/163

RPackages

R Packages

Studio


98/163

2014 RStudio, In

A collection of code and functions written

for the R language.

Usually focuses on a specific task orproblem.

Most of the useful R applications appear in

packages.

Start RStudio

Studio


99/163

2014 RStudio, In

ggplot2

Internet Your hard drive Your R session

Install your package withi ll k (" l 2")

Studio


100/163

ggplot2

2014 RStudio, In

install.packages("ggplot2")

ggplot2

Internet Your hard drive Your R session

install.packages("ggplot2")

Load your package withlib ( l t2)

( )EVERY

Studio


101/163

library(ggplot2)

2014 RStudio, InInternet Your hard drive Your R session

library(ggplot2)

ggplot2ggplot2

( )TIME

ggplot2

Studio


102/163

2014 RStudio, InInternet Your hard drive Your R session

library(ggplot2)

ggplot2ggplot2

ggplot2

qplot(1:10, 1:10)

Studio


103/163

2014 RStudio, In

## Error: could not find function "qplot"

You cannot use a function package until you load th

package

library(ggplot2)

qplot(1:10, 1:10)

##

Package summary

Studio


104/163

2014 RStudio, In

1.Download the package withinstall.packages("name")

% You only have to do this once

% You should be connected to the internet

2.Load the package withlibrary("name")

% You have to do this each time you start anR session.

There

areov

Rpa

c

Your Turn


105/163

2014 RStudio, In

We're going to use the ggplot2, maps,

RColorBrewer, and scales packages today.Load them with

library("ggplot2")

library("maps")

library("RColorBrewer")

Note: If you have not yet installed them, you'll need to install.packages(c("ggplot2", maps",

"RColorBrewer")) first.


106/163

Diamonds

Diamonds data


107/163

2014 RStudio, In

% ~54,000round diamonds from &

http://www.diamondse.info/

% comes in the ggplot2 package

% Carat, colour, clarity, cut

% Total depth, table, depth, &

width, height

% Price

Your turn
http://www.diamondse.info/


108/163

2014 RStudio, In

diamondsis huge!

Use subsetting to look at just the first sixrows of diamonds

Challenge: use subsetting to look at justthe lastsix rows

di d [1 6 ]

Studio


109/163

2014 RStudio, In

diamonds[1:6, ]

nrow(diamonds)

# 53940diamonds[53935:53940, ]

# Same as

head(diamonds)

tail(diamonds)

Studio

View


110/163

2014 RStudio, In

The View function can also help you examine a data

it opens a spreadsheet like data viewer.

View(diamonds)# notice: Capital V

Studio

Help pages


111/163

2014 RStudio, In

You can open the help page for any R object(including functions) by typing ?followed by theobject's name

?diamonds

table width

x


112/163

2014 RStudio, In

z

depth = z / diameter

table = table width / x * 100

x, y, z in mm


113/163

2014 RStudio, Inqplot(x, y, data = diamonds)

What is weirdabout these

values?


114/163


Can you removethem with


115/163


them withsubsetting?


116/163

Logicaltests

Studio

Logical comparisons


117/163

2014 RStudio, In

What will these return?

1 < 3

1 > 3

c(1, 2, 3, 4, 5) > 3

Your turnx


118/163

2014 RStudio, In

c( , , 3, , 5)

Your turnx


119/163

2014 RStudio, In

( , , , , )

Your turnx


120/163

2014 RStudio, In

( , , , , )

Your turnx


121/163

2014 RStudio, In

( )

Your turnx


122/163

2014 RStudio, In

Your turnx


123/163

2014 RStudio, In

Your turnx


124/163

2014 RStudio, In

%in%

Studio


125/163

2014 RStudio, In

# What does this do?

1 %in% c(1, 2, 3, 4)

1 %in% c(2, 3, 4)

c(3,4,5,6) %in% c(2, 3, 4)

%in%

Studio


126/163

2014 RStudio, In

%in%tests whether the object on the left is a

member of the group on the right.

1 %in% c(1, 2, 3, 4)

# TRUE

1 %in% c(2, 3, 4)

# FALSE

c(3,4,5,6) %in% c(2, 3, 4)

# TRUE TRUE FALSE FALSE

You can combine logical tests with & | xor ! any and

Studio

Boolean operators


127/163

2014 RStudio, In

You can combine logical tests with &, |, xor, !, any, andall

x > 2 & x < 9

TRUE & TRUE

TRUE


Studio

Boolean operators


128/163

2014 RStudio, In


x > 2 & x < 9

TRUE&TRUE

TRUE


Studio

Boolean operators


129/163

2014 RStudio, In


x > 2& x < 9

TRUE&TRUE

TRUE


Studio

Boolean operators


130/163

2014 RStudio, In


x > 2 & x < 9

TRUE & TRUE

TRUE

&

Are both condition 1 and condition 2 true?

Studio


131/163

2014 RStudio, In

Are both condition 1 andcondition 2 true?

|

Is either condition 1 or condition 2 true?

Studio


132/163

2014 RStudio, In

Is either condition 1 orcondition 2 true?

xor

Is either condition 1 or condition 2 true but not both?

Studio


133/163

2014 RStudio, In

Is either condition 1 orcondition 2 true, but not both?

!

Negation

Studio


134/163

2014 RStudio, In

Negation

any

Is any condition TRUE?

Studio


135/163

2014 RStudio, In

Is anycondition TRUE?

all

Is everycondition TRUE?

Studio


136/163

2014 RStudio, In

y

Studio

Logical operators


137/163

2014 RStudio, In

Studio

Boolean operators


138/163

2014 RStudio, In

Studio


139/163

2014 RStudio, In

w


140/163

2014 RStudio, In

Turn these sentences into logical tests in R

Is wpositive?

Is xgreater than 10 and less than 20?

Is object ythe word February?

Is every value in za day of the week?

# Answers

Studio


141/163

2014 RStudio, In

# Answers

w > 010 < x & x < 20

y == "February"

all(z %in% c("Monday", "Tuesday", "Wednesday

"Thursday", "Friday", "Saturday", "Sunday"

Studio


142/163

2014 RStudio, In

# Common mistakes

x > 10 & < 20

y = "February"

all(z == "Monday" | "Tuesday" | "Wednesday".

Studio

x > 10 & < 20??


143/163

2014 RStudio, In

x > 10 & < 20??

??TRUE & error!

error!

Studio

x > 10 & < 20??


144/163

2014 RStudio, In

x 10& 20??

??TRUE &error!

Studio

x > 10 & < 20??


145/163

2014 RStudio, In

x 10 & 20??

??TRUE &error!

error!

Studio

x > 10 & < 20??


146/163

2014 RStudio, In

??TRUE & error!

error!


147/163

Logicalsubsetting

Logical subsetting

Studio


148/163

2014 RStudio, In

Combining logical tests with subsetting is a verypowerful technique!

x_zeroes


149/163

2014 RStudio, In

# Prints to screen

diamonds[diamonds$x > 10, ]

# Saves to new data frame

big 10, ]

# Overwrites existing data frame.Dangerous!

diamonds


150/163

2014 RStudio, In

# Uh oh!

rm(diamonds)

str(diamonds)

# Phew!


151/163

Missingvalues

Studio

Data errors


152/163

2014 RStudio, In

Typically removing the entire row becauseof one error is overkill. Better to selectivelyreplace problem values with missingvalues.

In R, missing values are indicated by NA


153/163

2014 RStudio, In


154/163

2014 RStudio, In


155/163

2014 RStudio, In


156/163

2014 RStudio, In


157/163

2014 RStudio, In


158/163

2014 RStudio, In

Missing values propagate

Use i () to check for missing values

Studio

NA Behavior


159/163

2014 RStudio, In

Use is.na()to check for missing values

a


160/163

2014 RStudio, In

na.rmargument to remove missing values

prior to computation.

b


161/163

2014 RStudio, In

change individual values within an object

summary(diamonds$x)

diamonds$x[diamonds$x == 0]

diamonds$x[diamonds$x == 0]


162/163

2014 RStudio, In

diamonds$y[diamonds$y == 0]


163/163


03 the r language part 2

Documents