Preprocessing your data with R

Susan Holmes (c)

Missing data.
?is.na
example <- c("A",1,6,7,NA,"B")
example

## [1] "A" "1" "6" "7" NA "B"

mean(example)

## Warning in mean.default(example): argument is not numeric or logical:

## returning NA

## [1] NA

is.na(example)

## [1] FALSE FALSE FALSE FALSE TRUE FALSE

example2 <- c(2,1,6,7,NA,4)

example2

## [1] 2 1 6 7 NA 4

is.na(example2)

## [1] FALSE FALSE FALSE FALSE TRUE FALSE

mean(example2)

## [1] NA

length(example2)

## [1] 6

mean(example2,na.rm=TRUE)

## [1] 4

median(example2,na.rm=TRUE)

## [1] 4
Replacing just the missing values
example3 <- example2
example3

## [1] 2 1 6 7 NA 4

example3[is.na(example3)] <- 0
example3

## [1] 2 1 6 7 0 4

length(example3)

## [1] 6
Missing values can behave strangely
NA == NA

## [1] NA

NA+8

## [1] NA

NA^0

## [1] 1

1/NA

## [1] NA
Other strange values…
1/0

## [1] Inf

1/0-1/0

## [1] NaN
Data imputation
library("mice")
data(mammalsleep)
?mammalsleep
dim(mammalsleep)

## [1] 62 11

nic(mammalsleep)

## [1] 20

md.pattern(mammalsleep)

## species bw brw pi sei odi ts mls gt ps sws

## 42 1 1 1 1 1 1 1 1 1 1 1 0
## 2 1 1 1 1 1 1 1 0 1 1 1 1
## 3 1 1 1 1 1 1 1 1 0 1 1 1
## 9 1 1 1 1 1 1 1 1 1 0 0 2
## 2 1 1 1 1 1 1 0 1 1 1 0 2
## 1 1 1 1 1 1 1 1 0 0 1 1 2
## 2 1 1 1 1 1 1 0 1 1 0 0 3
## 1 1 1 1 1 1 1 1 0 1 0 0 3
## 0 0 0 0 0 0 4 4 4 12 14 38

Missing at random (MCAR) versus systematic patterns (MNAR).

?mice
Outlier detection
summary(mammalsleep)

## species bw brw
## African elephant : 1 Min. : 0.005 Min. : 0.14
## African giant pouched rat: 1 1st Qu.: 0.600 1st Qu.: 4.25
## Arctic Fox : 1 Median : 3.342 Median : 17.25
## Arctic ground squirrel : 1 Mean : 198.790 Mean : 283.13
## Asian elephant : 1 3rd Qu.: 48.203 3rd Qu.: 166.00
## Baboon : 1 Max. :6654.000 Max. :5712.00
## (Other) :56
## sws ps ts mls
## Min. : 2.100 Min. :0.000 Min. : 2.60 Min. : 2.000
## 1st Qu.: 6.250 1st Qu.:0.900 1st Qu.: 8.05 1st Qu.: 6.625
## Median : 8.350 Median :1.800 Median :10.45 Median : 15.100
## Mean : 8.673 Mean :1.972 Mean :10.53 Mean : 19.878
## 3rd Qu.:11.000 3rd Qu.:2.550 3rd Qu.:13.20 3rd Qu.: 27.750
## Max. :17.900 Max. :6.600 Max. :19.90 Max. :100.000
## NA's :14 NA's :12 NA's :4 NA's :4
## gt pi sei odi
## Min. : 12.00 Min. :1.000 Min. :1.000 Min. :1.000
## 1st Qu.: 35.75 1st Qu.:2.000 1st Qu.:1.000 1st Qu.:1.000
## Median : 79.00 Median :3.000 Median :2.000 Median :2.000
## Mean :142.35 Mean :2.871 Mean :2.419 Mean :2.613
## 3rd Qu.:207.50 3rd Qu.:4.000 3rd Qu.:4.000 3rd Qu.:4.000
## Max. :645.00 Max. :5.000 Max. :5.000 Max. :5.000
## NA's :4

which.max(mammalsleep$bw)

## [1] 1

mammalsleep[which.max(mammalsleep$bw),]

## species bw brw sws ps ts mls gt pi sei odi

## 1 African elephant 6654 5712 NA NA 3.3 38.6 645 3 5 3

Document them, find the reason they occurred, then remove them.
Make the data easier to look at
interactively
View(pressure)
View(iris)
Grouping Data
load('births.RData')
head(births)

## year month date_of_month day_of_week births

## 1 2000 1 1 6 9083
## 2 2000 1 2 7 8006
## 3 2000 1 3 1 11363
## 4 2000 1 4 2 13032
## 5 2000 1 5 3 12558
## 6 2000 1 6 4 12466

birthn <- births

save(birthn,file="birthn.RData")
Different ways of filtering the data
Choosing only the Saturday births.

###Subsetting
Sat <-birthn[birthn$day_of_week==6,]
Sat[1:5,]

## year month date_of_month day_of_week births

## 1 2000 1 1 6 9083
## 8 2000 1 8 6 8934
## 15 2000 1 15 6 8525
## 22 2000 1 22 6 8855
## 29 2000 1 29 6 8805

library(dplyr)

##
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':

##
## filter, lag

## The following objects are masked from 'package:base':

##
## intersect, setdiff, setequal, union

Sat1 <-filter(birthn, day_of_week == 6)

Sat1[1:5,]

## year month date_of_month day_of_week births

## 1 2000 1 1 6 9083
## 2 2000 1 8 6 8934
## 3 2000 1 15 6 8525
## 4 2000 1 22 6 8855
## 5 2000 1 29 6 8805

Sat2 <- birthn %>% filter(day_of_week == 6)

Sat2[1:5,]

## year month date_of_month day_of_week births

## 1 2000 1 1 6 9083
## 2 2000 1 8 6 8934
## 3 2000 1 15 6 8525
## 4 2000 1 22 6 8855
## 5 2000 1 29 6 8805

Another way of looking at data is to make them into what is called a tibble: (tbl:
tibble).

tbl s have the advantage of always showing themselves in the console optimally.
tbl_df gives similar information as str we have been using.

tbl_df(Sat1)

## # A tibble: 783 × 5
## year month date_of_month day_of_week births
## <int> <int> <int> <int> <int>
## 1 2000 1 1 6 9083
## 2 2000 1 8 6 8934
## 3 2000 1 15 6 8525
## 4 2000 1 22 6 8855
## 5 2000 1 29 6 8805
## 6 2000 2 5 6 8624
## 7 2000 2 12 6 8836
## 8 2000 2 19 6 8861
## 9 2000 2 26 6 9026
## 10 2000 3 4 6 9054
## # ... with 773 more rows

str(Sat1)

## 'data.frame': 783 obs. of 5 variables:

## $ year : int 2000 2000 2000 2000 2000 2000 2000 2000 2000 2000 ...
## $ month : int 1 1 1 1 1 2 2 2 2 3 ...
## $ date_of_month: int 1 8 15 22 29 5 12 19 26 4 ...
## $ day_of_week : int 6 6 6 6 6 6 6 6 6 6 ...
## $ births : int 9083 8934 8525 8855 8805 8624 8836 8861 9026 9054 ...
Sequences of Transformations
The %>% operator helps when we are doing several nested operations.

Here is an example

GroupBirths <- group_by(birthn,day_of_week)

GroupMeans <- summarise(GroupBirths,mean(births))
SortedBirths <- arrange(GroupMeans, `mean(births)`)
SortedBirths

## # A tibble: 7 × 2
## day_of_week `mean(births)`
## <int> <dbl>
## 1 7 7518.377
## 2 6 8562.573
## 3 1 11897.830
## 4 5 12596.162
## 5 4 12845.826
## 6 3 12910.766
## 7 2 13122.444

str(SortedBirths)

## Classes 'tbl_df', 'tbl' and 'data.frame': 7 obs. of 2 variables:

## $ day_of_week : int 7 6 1 5 4 3 2
## $ mean(births): num 7518 8563 11898 12596 12846 ...

birthn %>%
group_by(day_of_week) %>%
summarise(avg = mean(births)) %>%
arrange(avg)

## # A tibble: 7 × 2
## day_of_week avg
## <int> <dbl>
## 1 7 7518.377
## 2 6 8562.573
## 3 1 11897.830
## 4 5 12596.162
## 5 4 12845.826
## 6 3 12910.766
## 7 2 13122.444

####More succintly
birthn %>%
group_by(day_of_week) %>%
summarise(mean(births)) %>%
arrange()

## # A tibble: 7 × 2
## day_of_week `mean(births)`
## <int> <dbl>
## 1 1 11897.830
## 2 2 13122.444
 ## 3 3 12910.766
## 4 4 12845.826
## 5 5 12596.162
## 6 6 8562.573
## 7 7 7518.377

x %>% f(y) is equivalent to just executing f(x,y)

If we need to execute a sequence of functions: h(g(f(x,y),z),m) can be hard to parse

and read.

x %>% f(y) %>% g(z) %>% h(m) gives the same answer.

To find out the average of Friday 13th births:

birthn %>%
filter(day_of_week == 5) %>%
filter(date_of_month == 13) %>%
summarise(mean(births))

## mean(births)
## 1 11949.96

birthn %>%
filter(day_of_week < 5) %>%
filter(date_of_month != 13) %>%
summarise(mean(births))

## mean(births)
## 1 12700.61
Bad Drivers Data
Five Thirty Eight Article

You need an internet connection for this to work:

drivers <- read.csv(url("https://raw.githubusercontent.com/fivethirtyeight/data/master/bad-drivers/bad-drivers.csv"))

head(drivers)

## State
## 1 Alabama
## 2 Alaska
## 3 Arizona
## 4 Arkansas
## 5 California
## 6 Colorado
## Number.of.drivers.involved.in.fatal.collisions.per.billion.miles
## 1 18.8
## 2 18.1
## 3 18.6
## 4 22.4
## 5 12.0
## 6 13.6
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Speeding
## 1 39
## 2 41
## 3 35
## 4 18
## 5 35
## 6 37
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Alcohol.Impaired
## 1 30
## 2 25
## 3 28
## 4 26
## 5 28
## 6 28
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Not.Distracted
## 1 96
## 2 90
## 3 84
## 4 94
## 5 91
## 6 79
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Had.Not.Been.Involved.In.Any.Previous.Accidents
## 1 80
## 2 94
## 3 96
## 4 95
## 5 89
## 6 95
## Car.Insurance.Premiums....
## 1 784.55
## 2 1053.48
## 3 899.47
## 4 827.34
## 5 878.41
## 6 835.50
## Losses.incurred.by.insurance.companies.for.collisions.per.insured.driver....
## 1 145.08
## 2 133.93
## 3 110.35
## 4 142.39
## 5 165.63
## 6 139.91

tbl_df(drivers)

## # A tibble: 51 × 8
## State
## <fctr>
## 1 Alabama
## 2 Alaska
## 3 Arizona
## 4 Arkansas
## 5 California
## 6 Colorado
## 7 Connecticut
## 8 Delaware
## 9 District of Columbia
## 10 Florida
## # ... with 41 more rows, and 7 more variables:
## # Number.of.drivers.involved.in.fatal.collisions.per.billion.miles <dbl>,
## # Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Speeding <int>,
## # Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Alcohol.Impaired <int>,
## # Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Not.Distracted <int>,
## # Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Had.Not.Been.Involved.In.Any.Previous.Accident
## # Car.Insurance.Premiums.... <dbl>,
## # Losses.incurred.by.insurance.companies.for.collisions.per.insured.driver.... <dbl>

glimpse(drivers)

## Observations: 51
## Variables: 8
## $ State <
## $ Number.of.drivers.involved.in.fatal.collisions.per.billion.miles <
## $ Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Speeding <
## $ Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Alcohol.Impaired <
## $ Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Not.Distracted <
## $ Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Had.Not.Been.Involved.In.Any.Previous.Accidents <
## $ Car.Insurance.Premiums.... <
## $ Losses.incurred.by.insurance.companies.for.collisions.per.insured.driver.... <

summary(drivers)

## State
## Alabama : 1
## Alaska : 1
## Arizona : 1
## Arkansas : 1
## California: 1
## Colorado : 1
## (Other) :45
## Number.of.drivers.involved.in.fatal.collisions.per.billion.miles
## Min. : 5.90
## 1st Qu.:12.75
## Median :15.60
## Mean :15.79
## 3rd Qu.:18.50
## Max. :23.90
##
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Speeding
## Min. :13.00
## 1st Qu.:23.00
## Median :34.00
## Mean :31.73
## 3rd Qu.:38.00
## Max. :54.00
##
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Alcohol.Impaired
## Min. :16.00
## 1st Qu.:28.00
## Median :30.00
## Mean :30.69
## 3rd Qu.:33.00
## Max. :44.00
##
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Were.Not.Distracted
## Min. : 10.00
## 1st Qu.: 83.00
## Median : 88.00
## Mean : 85.92
## 3rd Qu.: 95.00
## Max. :100.00
##
## Percentage.Of.Drivers.Involved.In.Fatal.Collisions.Who.Had.Not.Been.Involved.In.Any.Previous.Accidents
## Min. : 76.00
## 1st Qu.: 83.50
## Median : 88.00
## Mean : 88.73
## 3rd Qu.: 95.00
## Max. :100.00
##
## Car.Insurance.Premiums....
## Min. : 642.0
## 1st Qu.: 768.4
## Median : 859.0
## Mean : 887.0
## 3rd Qu.:1007.9
## Max. :1301.5
##
## Losses.incurred.by.insurance.companies.for.collisions.per.insured.driver....
## Min. : 82.75
## 1st Qu.:114.64
## Median :136.05
## Mean :134.49
## 3rd Qu.:151.87
## Max. :194.78
##

colnames(drivers)=
c("State","NperB","PrcSpeed","PrcAlco","PrcNotDist","PrcNoPrev","Premium","Loss")
sort(drivers[,2])

## [1] 5.9 8.2 9.6 10.6 10.8 11.1 11.2 11.3 11.6 12.0 12.3 12.5 12.7 12.8
## [15] 12.8 13.6 13.6 13.8 14.1 14.1 14.5 14.7 14.9 15.1 15.3 15.6 15.7 16.1
## [29] 16.2 16.8 17.4 17.5 17.6 17.8 17.9 18.1 18.2 18.4 18.6 18.8 19.4 19.4
## [43] 19.5 19.9 20.5 21.4 21.4 22.4 23.8 23.9 23.9

drivers[1:10,1:3]

## State NperB PrcSpeed

## 1 Alabama 18.8 39
## 2 Alaska 18.1 41
## 3 Arizona 18.6 35
## 4 Arkansas 22.4 18
## 5 California 12.0 35
## 6 Colorado 13.6 37
## 7 Connecticut 10.8 46
## 8 Delaware 16.2 38
## 9 District of Columbia 5.9 34
## 10 Florida 17.9 21

drivers[order(drivers[,2]),1:3]

## State NperB PrcSpeed

## 9 District of Columbia 5.9 34
## 22 Massachusetts 8.2 23
## 24 Minnesota 9.6 23
## 48 Washington 10.6 42
## 7 Connecticut 10.8 46
## 40 Rhode Island 11.1 34
## 31 New Jersey 11.2 16
## 45 Utah 11.3 43
## 30 New Hampshire 11.6 35
## 5 California 12.0 35
## 33 New York 12.3 32
## 21 Maryland 12.5 34
## 47 Virginia 12.7 19
## 14 Illinois 12.8 36
## 38 Oregon 12.8 33
## 6 Colorado 13.6 37
## 46 Vermont 13.6 30
## 50 Wisconsin 13.8 36
## 23 Michigan 14.1 24
## 36 Ohio 14.1 28
## 15 Indiana 14.5 25
## 29 Nevada 14.7 37
## 28 Nebraska 14.9 13
## 20 Maine 15.1 38
## 13 Idaho 15.3 36
## 11 Georgia 15.6 19
## 16 Iowa 15.7 17
## 26 Missouri 16.1 43
## 8 Delaware 16.2 38
## 34 North Carolina 16.8 39
## 51 Wyoming 17.4 42
## 12 Hawaii 17.5 54
## 25 Mississippi 17.6 15
## 17 Kansas 17.8 27
## 10 Florida 17.9 21
## 2 Alaska 18.1 41
## 39 Pennsylvania 18.2 50
## 32 New Mexico 18.4 19
## 3 Arizona 18.6 35
## 1 Alabama 18.8 39
## 42 South Dakota 19.4 31
## 44 Texas 19.4 40
## 43 Tennessee 19.5 21
## 37 Oklahoma 19.9 32
## 19 Louisiana 20.5 35
## 18 Kentucky 21.4 19
## 27 Montana 21.4 39
## 4 Arkansas 22.4 18
## 49 West Virginia 23.8 34
## 35 North Dakota 23.9 23
## 41 South Carolina 23.9 38

arrange(drivers,NperB)

## State NperB PrcSpeed PrcAlco PrcNotDist PrcNoPrev

## 1 District of Columbia 5.9 34 27 100 100
## 2 Massachusetts 8.2 23 35 87 80
## 3 Minnesota 9.6 23 29 88 88
## 4 Washington 10.6 42 33 82 86
## 5 Connecticut 10.8 46 36 87 82
## 6 Rhode Island 11.1 34 38 92 79
## 7 New Jersey 11.2 16 28 86 78
## 8 Utah 11.3 43 16 88 96
## 9 New Hampshire 11.6 35 30 87 83
## 10 California 12.0 35 28 91 89
## 11 New York 12.3 32 29 88 80
## 12 Maryland 12.5 34 32 71 99
## 13 Virginia 12.7 19 27 87 88
## 14 Illinois 12.8 36 34 94 96
## 15 Oregon 12.8 33 26 67 90
## 16 Colorado 13.6 37 28 79 95
## 17 Vermont 13.6 30 30 96 95
## 18 Wisconsin 13.8 36 33 39 84
## 19 Michigan 14.1 24 28 95 77
## 20 Ohio 14.1 28 34 99 82
## 21 Indiana 14.5 25 29 95 95
## 22 Nevada 14.7 37 32 95 99
## 23 Nebraska 14.9 13 35 93 90
## 24 Maine 15.1 38 30 87 84
## 25 Idaho 15.3 36 29 85 98
## 26 Georgia 15.6 19 25 95 93
## 27 Iowa 15.7 17 25 97 87
## 28 Missouri 16.1 43 34 92 84
## 29 Delaware 16.2 38 30 87 99
## 30 North Carolina 16.8 39 31 94 81
## 31 Wyoming 17.4 42 32 81 90
## 32 Hawaii 17.5 54 41 82 87
## 33 Mississippi 17.6 15 31 10 100
## 34 Kansas 17.8 27 24 77 85
## 35 Florida 17.9 21 29 92 94
## 36 Alaska 18.1 41 25 90 94
## 37 Pennsylvania 18.2 50 31 96 88
## 38 New Mexico 18.4 19 27 67 98
## 39 Arizona 18.6 35 28 84 96
## 40 Alabama 18.8 39 30 96 80
## 41 South Dakota 19.4 31 33 98 86
## 42 Texas 19.4 40 38 91 87
## 43 Tennessee 19.5 21 29 82 81
## 44 Oklahoma 19.9 32 29 92 94
## 45 Louisiana 20.5 35 33 73 98
## 46 Kentucky 21.4 19 23 78 76
## 47 Montana 21.4 39 44 84 85
## 48 Arkansas 22.4 18 26 94 95
## 49 West Virginia 23.8 34 28 97 87
## 50 North Dakota 23.9 23 42 99 86
## 51 South Carolina 23.9 38 41 96 81
## Premium Loss
## 1 1273.89 136.05
## 2 1011.14 135.63
## 3 777.18 133.35
## 4 890.03 111.62
## 5 1068.73 167.02
## 6 1148.99 148.58
## 7 1301.52 159.85
## 8 809.38 109.48
## 9 746.54 120.21
## 10 878.41 165.63
## 11 1234.31 150.01
## 12 1048.78 192.70
## 13 768.95 153.72
## 14 803.11 139.15
## 15 804.71 104.61
## 16 835.50 139.91
## 17 716.20 109.61
## 18 670.31 106.62
## 19 1110.61 152.26
## 20 697.73 133.52
## 21 710.46 108.92
## 22 1029.87 138.71
## 23 732.28 114.82
## 24 661.88 96.57
## 25 641.96 82.75
## 26 913.15 142.80
## 27 649.06 114.47
## 28 790.32 144.45
## 29 1137.87 151.48
## 30 708.24 127.82
## 31 791.14 122.04
## 32 861.18 120.92
## 33 896.07 155.77
## 34 780.45 133.80
## 35 1160.13 144.18
## 36 1053.48 133.93
## 37 905.99 153.86
## 38 869.85 120.75
## 39 899.47 110.35
## 40 784.55 145.08
## 41 669.31 96.87
## 42 1004.75 156.83
## 43 767.91 155.57
## 44 881.51 178.86
## 45 1281.55 194.78
## 46 872.51 137.13
## 47 816.21 85.15
## 48 827.34 142.39
## 49 992.61 152.56
## 50 688.75 109.72
## 51 858.97 116.29

arrange(drivers,desc(PrcSpeed))

## State NperB PrcSpeed PrcAlco PrcNotDist PrcNoPrev

## 1 Hawaii 17.5 54 41 82 87
## 2 Pennsylvania 18.2 50 31 96 88
## 3 Connecticut 10.8 46 36 87 82
## 4 Missouri 16.1 43 34 92 84
## 5 Utah 11.3 43 16 88 96
## 6 Washington 10.6 42 33 82 86
## 7 Wyoming 17.4 42 32 81 90
## 8 Alaska 18.1 41 25 90 94
## 9 Texas 19.4 40 38 91 87
## 10 Alabama 18.8 39 30 96 80
## 11 Montana 21.4 39 44 84 85
## 12 North Carolina 16.8 39 31 94 81
## 13 Delaware 16.2 38 30 87 99
## 14 Maine 15.1 38 30 87 84
## 15 South Carolina 23.9 38 41 96 81
## 16 Colorado 13.6 37 28 79 95
## 17 Nevada 14.7 37 32 95 99
## 18 Idaho 15.3 36 29 85 98
## 19 Illinois 12.8 36 34 94 96
## 20 Wisconsin 13.8 36 33 39 84
## 21 Arizona 18.6 35 28 84 96
## 22 California 12.0 35 28 91 89
## 23 Louisiana 20.5 35 33 73 98
## 24 New Hampshire 11.6 35 30 87 83
## 25 District of Columbia 5.9 34 27 100 100
## 26 Maryland 12.5 34 32 71 99
## 27 Rhode Island 11.1 34 38 92 79
## 28 West Virginia 23.8 34 28 97 87
## 29 Oregon 12.8 33 26 67 90
## 30 New York 12.3 32 29 88 80
## 31 Oklahoma 19.9 32 29 92 94
## 32 South Dakota 19.4 31 33 98 86
## 33 Vermont 13.6 30 30 96 95
## 34 Ohio 14.1 28 34 99 82
## 35 Kansas 17.8 27 24 77 85
## 36 Indiana 14.5 25 29 95 95
## 37 Michigan 14.1 24 28 95 77
## 38 Massachusetts 8.2 23 35 87 80
## 39 Minnesota 9.6 23 29 88 88
## 40 North Dakota 23.9 23 42 99 86
## 41 Florida 17.9 21 29 92 94
## 42 Tennessee 19.5 21 29 82 81
## 43 Georgia 15.6 19 25 95 93
## 44 Kentucky 21.4 19 23 78 76
## 45 New Mexico 18.4 19 27 67 98
## 46 Virginia 12.7 19 27 87 88
## 47 Arkansas 22.4 18 26 94 95
## 48 Iowa 15.7 17 25 97 87
## 49 New Jersey 11.2 16 28 86 78
## 50 Mississippi 17.6 15 31 10 100
## 51 Nebraska 14.9 13 35 93 90
## Premium Loss
## 1 861.18 120.92
## 2 905.99 153.86
## 3 1068.73 167.02
## 4 790.32 144.45
## 5 809.38 109.48
## 6 890.03 111.62
## 7 791.14 122.04
## 8 1053.48 133.93
## 9 1004.75 156.83
## 10 784.55 145.08
## 11 816.21 85.15
## 12 708.24 127.82
## 13 1137.87 151.48
## 14 661.88 96.57
## 15 858.97 116.29
## 16 835.50 139.91
## 17 1029.87 138.71
## 18 641.96 82.75
## 19 803.11 139.15
## 20 670.31 106.62
## 21 899.47 110.35
## 22 878.41 165.63
## 23 1281.55 194.78
## 24 746.54 120.21
## 25 1273.89 136.05
## 26 1048.78 192.70
## 27 1148.99 148.58
## 28 992.61 152.56
## 29 804.71 104.61
## 30 1234.31 150.01
## 31 881.51 178.86
## 32 669.31 96.87
## 33 716.20 109.61
## 34 697.73 133.52
## 35 780.45 133.80
## 36 710.46 108.92
## 37 1110.61 152.26
## 38 1011.14 135.63
## 39 777.18 133.35
## 40 688.75 109.72
## 41 1160.13 144.18
## 42 767.91 155.57
## 43 913.15 142.80
## 44 872.51 137.13
## 45 869.85 120.75
## 46 768.95 153.72
## 47 827.34 142.39
## 48 649.06 114.47
## 49 1301.52 159.85
## 50 896.07 155.77
## 51 732.28 114.82
Make new variables

driversp=mutate(drivers,prem_c=Loss/Premium)
select(arrange(driversp,prem_c),State,prem_c)

## State prem_c
## 1 Montana 0.1043236
## 2 District of Columbia 0.1067989
## 3 New York 0.1215335
## 4 Arizona 0.1226834
## 5 New Jersey 0.1228179
## 6 Florida 0.1242792
## 7 Washington 0.1254115
## 8 Alaska 0.1271310
## 9 Idaho 0.1289021
## 10 Rhode Island 0.1293136
## 11 Oregon 0.1299971
## 12 Delaware 0.1331259
## 13 Massachusetts 0.1341357
## 14 Nevada 0.1346869
## 15 Utah 0.1352640
## 16 South Carolina 0.1353831
## 17 Michigan 0.1370958
## 18 New Mexico 0.1388170
## 19 Hawaii 0.1404120
## 20 South Dakota 0.1447311
## 21 Maine 0.1459026
## 22 Louisiana 0.1519878
## 23 Vermont 0.1530438
## 24 Indiana 0.1533091
## 25 West Virginia 0.1536958
## 26 Wyoming 0.1542584
## 27 Texas 0.1560886
## 28 Connecticut 0.1562789
## 29 Georgia 0.1563818
## 30 Nebraska 0.1567979
## 31 Kentucky 0.1571673
## 32 Wisconsin 0.1590607
## 33 North Dakota 0.1593031
## 34 New Hampshire 0.1610229
## 35 Colorado 0.1674566
## 36 Pennsylvania 0.1698253
## 37 Kansas 0.1714396
## 38 Minnesota 0.1715819
## 39 Arkansas 0.1721058
## 40 Illinois 0.1732639
## 41 Mississippi 0.1738369
## 42 Iowa 0.1763627
## 43 North Carolina 0.1804755
## 44 Missouri 0.1827741
## 45 Maryland 0.1837373
## 46 Alabama 0.1849213
## 47 California 0.1885566
## 48 Ohio 0.1913634
## 49 Virginia 0.1999090
## 50 Tennessee 0.2025888
## 51 Oklahoma 0.2029018
Document all the changes you make using a
script.
The best way to make a report is to put everything into an .Rmd document and
then knit into an html file using the knitr package.
Summary of this Session:
Careful data preprocessing is necessary at the beginning of any data exploration
exercise.

Missing data and outliers need to be identified.

Missing data may be imputed if there are only a few in a column or row and if
their occurrence patterns are random.

We saw how to use the package dplyr that allows us to easily do a sequence of
actions on data using the %>% operator.

Some of the possible actions are:

filter()
arrange()
select()
mutate()
summarise()
sample_n()

We saw that preprocessing your data should be documented with scripts

you save. A good way to do this is to use RStudio’s Rmd editor and html
generator.

Question: Look up the RStudio data wrangling cheatsheet: R Data Wrangling

Cheatsheet

Activity: Re-analyze the drivers data and make your own Rmd and html reports.