Basic use of R for reading, manipulating, and plotting data!

• Understand types of data and their distributions

• Numerical summaries

• Understand types of data and their distributions

• Numerical summaries (across subgroups)

• Understand types of data and their distributions

• Numerical summaries (across subgroups)

  • Contingency Tables
    
  • Mean/Median
    
  • Standard Deviation/Variance/IQR
  • Quantiles/Percentiles

• Understand types of data and their distributions

• Numerical summaries (across subgroups)

  • Contingency Tables
    
  • Mean/Median
    
  • Standard Deviation/Variance/IQR
  • Quantiles/Percentiles

• Graphical summaries (across subgroups)

  • Bar plots
    
  • Histograms
    
  • Box plots
    
  • Scatter plots

• How to summarize data?

• Depends on data type:

  • Categorical (Qualitative) variable - entries are a label or attribute
    
  • Numeric (Quantitative) variable - entries are a numerical value where math can be performed

Common goal: Describe the distribution of the variable

• Distribution = pattern and frequency with which you observe a variable

• Categorical variable - entries are a label or attribute

• Try to describe relative frequency (or count) in each category

• Create contingency tables via table

• Consider data on titanic passengers in titanic.csv

titanicData <- read_csv("../datasets/titanic.csv")
titanicData

## # A tibble: 1,310 x 14
##   pclass survived name      sex      age sibsp parch ticket  fare cabin embarked
##    <dbl>    <dbl> <chr>     <chr>  <dbl> <dbl> <dbl> <chr>  <dbl> <chr> <chr>   
## 1      1        1 Allen, M~ fema~ 29         0     0 24160   211. B5    S       
## 2      1        1 Allison,~ male   0.917     1     2 113781  152. C22 ~ S       
## 3      1        0 Allison,~ fema~  2         1     2 113781  152. C22 ~ S       
## 4      1        0 Allison,~ male  30         1     2 113781  152. C22 ~ S       
## 5      1        0 Allison,~ fema~ 25         1     2 113781  152. C22 ~ S       
## # ... with 1,305 more rows, and 3 more variables: boat <chr>, body <dbl>,
## #   home.dest <chr>

• Create one-way contingency tables for each of three categorical variables:

  • embarked (where journey started)
    
  • survived (survive or not)
    
  • sex (Male or Female)

table(titanicData$embarked)

## 
##   C   Q   S 
## 270 123 914

table(titanicData$survived)

## 
##   0   1 
## 809 500

table(titanicData$sex)

## 
## female   male 
##    466    843

• Create two-way contingency tables for pairs of categorical variables

table(titanicData$survived,
      titanicData$sex)

##    
##     female male
##   0    127  682
##   1    339  161

table(titanicData$survived,
      titanicData$embarked)

##    
##       C   Q   S
##   0 120  79 610
##   1 150  44 304

table(titanicData$sex,
      titanicData$embarked)

##         
##            C   Q   S
##   female 113  60 291
##   male   157  63 623

• Create a three-way contingency table for three categorical variables

table(titanicData$sex, titanicData$embarked, titanicData$survived)

## , ,  = 0
## 
##         
##            C   Q   S
##   female  11  23  93
##   male   109  56 517
## 
## , ,  = 1
## 
##         
##            C   Q   S
##   female 102  37 198
##   male    48   7 106

• Create a three-way contingency table for three categorical variables (order matters for output!)

• Example of an array! 3 dimensions [ , , ]

tab <- table(titanicData$sex, titanicData$embarked, titanicData$survived)

str(tab)

##  'table' int [1:2, 1:3, 1:2] 11 109 23 56 93 517 102 48 37 7 ...
##  - attr(*, "dimnames")=List of 3
##   ..$ : chr [1:2] "female" "male"
##   ..$ : chr [1:3] "C" "Q" "S"
##   ..$ : chr [1:2] "0" "1"

• Can obtain conditional bivariate info!

##  'table' int [1:2, 1:3, 1:2] 11 109 23 56 93 517 102 48 37 7 ...
##  - attr(*, "dimnames")=List of 3
##   ..$ : chr [1:2] "female" "male"
##   ..$ : chr [1:3] "C" "Q" "S"
##   ..$ : chr [1:2] "0" "1"

#returns embarked vs survived table for females
tab[1, , ]

##    
##       0   1
##   C  11 102
##   Q  23  37
##   S  93 198

• Can obtain conditional bivariate info!

##  'table' int [1:2, 1:3, 1:2] 11 109 23 56 93 517 102 48 37 7 ...
##  - attr(*, "dimnames")=List of 3
##   ..$ : chr [1:2] "female" "male"
##   ..$ : chr [1:3] "C" "Q" "S"
##   ..$ : chr [1:2] "0" "1"

#returns embarked vs survived table for males
tab[2, , ]

##    
##       0   1
##   C 109  48
##   Q  56   7
##   S 517 106

• Can obtain conditional bivariate info!

##  'table' int [1:2, 1:3, 1:2] 11 109 23 56 93 517 102 48 37 7 ...
##  - attr(*, "dimnames")=List of 3
##   ..$ : chr [1:2] "female" "male"
##   ..$ : chr [1:3] "C" "Q" "S"
##   ..$ : chr [1:2] "0" "1"

#returns survived vs sex table for embarked "C"
tab[, 1, ]

##         
##            0   1
##   female  11 102
##   male   109  48

• Can obtain conditional univariate info too!

##  'table' int [1:2, 1:3, 1:2] 11 109 23 56 93 517 102 48 37 7 ...
##  - attr(*, "dimnames")=List of 3
##   ..$ : chr [1:2] "female" "male"
##   ..$ : chr [1:3] "C" "Q" "S"
##   ..$ : chr [1:2] "0" "1"

#Survived status for males that embarked at "Q"
tab[2, 2, ]

##  0  1 
## 56  7

Numeric variable - entries are a numerical value where math can be performed

Single variable: describe the distribution via

• Shape: Histogram, Density plot, …

• Measures of center: Mean, Median, …

• Measures of spread: Variance, Standard Deviation, Quartiles, IQR, …

Numeric variable - entries are a numerical value where math can be performed

Single variable: describe the distribution via

• Shape: Histogram, Density plot, …

• Measures of center: Mean, Median, …

• Measures of spread: Variance, Standard Deviation, Quartiles, IQR, …

Two Variables:

• Shape: Scatter plot, …

• Measures of linear relationship: Covariance, Correlation, …

• Look at carbon dioxide (CO2) uptake data set

  • Response recorded: uptake CO2 uptake rates in grass plants
    
  • Environment manipulated: Treatment - chilled/nonchilled
    
  • Ambient CO2 specified and measured: conc

CO2 <- as_tibble(CO2)
CO2

## # A tibble: 84 x 5
##   Plant Type   Treatment   conc uptake
##   <ord> <fct>  <fct>      <dbl>  <dbl>
## 1 Qn1   Quebec nonchilled    95   16  
## 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
## # ... with 79 more rows

Mean & Median

mean(CO2$uptake)

## [1] 27.2131

#note you can easily get a trimmed mean
mean(CO2$uptake, trim = 0.05) #5% trimmed mean

## [1] 27.25263

median(CO2$uptake)

## [1] 28.3

Variance, Standard Deviation, Quartiles, & IQR

#quartiles and mean
summary(CO2$uptake)

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##    7.70   17.90   28.30   27.21   37.12   45.50

var(CO2$uptake)

## [1] 116.9515

sd(CO2$uptake)

## [1] 10.81441

IQR(CO2$uptake)

## [1] 19.225

quantile(CO2$uptake, probs = c(0.1, 0.2))

##   10%   20% 
## 12.36 15.64

Covariance & Correlation

cov(CO2$conc, CO2$uptake)

## [1] 1552.687

cor(CO2$conc, CO2$uptake)

## [1] 0.4851774

Usually want summaries for different subgroups of data

• Ex: Get similar uptake summaries for each Treatment

Usually want summaries for different subgroups of data

• Ex: Get similar uptake summaries for each Treatment

• dplyr easy to use but can only return one value

Usually want summaries for different subgroups of data

• Ex: Get similar uptake summaries for each Treatment

• dplyr easy to use (although it can only return one value)

Idea:

• Use group_by to create subgroups associated with the data frame

• Use summarize to create basic summaries for each subgroup

• Ex: Get similar uptake summaries for each Treatment

CO2 %>% 
  group_by(Treatment) %>% 
    summarise(avg = mean(uptake), med = median(uptake), var = var(uptake))

## # A tibble: 2 x 4
##   Treatment    avg   med   var
##   <fct>      <dbl> <dbl> <dbl>
## 1 nonchilled  30.6  31.3  94.2
## 2 chilled     23.8  19.7 118.

• Ex: Get similar uptake summaries for each Treatment and Concentration

CO2 %>% 
  group_by(Treatment, conc) %>% 
        summarise(avg = mean(uptake), med = median(uptake), var = var(uptake))

## # A tibble: 14 x 5
## # Groups:   Treatment [2]
##    Treatment   conc   avg   med    var
##    <fct>      <dbl> <dbl> <dbl>  <dbl>
##  1 nonchilled    95  13.3  12.8   5.75
##  2 nonchilled   175  25.1  24.6  32.6 
##  3 nonchilled   250  32.5  32.7  35.1 
##  4 nonchilled   350  35.1  34.5  37.4 
##  5 nonchilled   500  35.1  33.8  31.9 
##  6 nonchilled   675  36.0  35.8  40.2 
##  7 nonchilled  1000  37.4  37.6  49.8 
##  8 chilled       95  11.2  10.6   8.18
##  9 chilled      175  19.4  19.5  34.7 
## 10 chilled      250  25.3  24.2 112.  
## 11 chilled      350  26.2  26.4 117.  
## 12 chilled      500  26.6  26   131.  
## 13 chilled      675  27.9  28.8 120.  
## 14 chilled     1000  29.8  30.3 154.

dplyr has variations on summarise that can be used:

• summarise_all() - Apply functions to every column

• summarise_at() - Apply functions to specific columns

• summarise_if() - Apply functions to all columns of one type

• Ex: Get similar uptake summaries for each Treatment

• Built-in aggregate() function more general

• Ex: Get similar uptake summaries for each Treatment

• Built-in aggregate() function more general

• Basic use gives response (x) and a list of variables to group by

aggregate(x = CO2$uptake, by = list(CO2$Treatment), FUN = summary)

##      Group.1   x.Min. x.1st Qu. x.Median   x.Mean x.3rd Qu.   x.Max.
## 1 nonchilled 10.60000  26.47500 31.30000 30.64286  38.70000 45.50000
## 2    chilled  7.70000  14.52500 19.70000 23.78333  34.90000 42.40000

• aggregate() is commonly used with formula notation!

uptake ~ Treatment - is an example of formula notation

• Idea: uptake (LHS) modeled by Treatment levels (RHS)

• aggregate() is commonly used with formula notation!

uptake ~ Treatment - is an example of formula notation

• Idea: uptake (LHS) modeled by Treatment levels (RHS)

aggregate(uptake ~ Treatment, data = CO2, FUN = summary)

• aggregate() is commonly used with formula notation!

uptake ~ Treatment + conc model uptake by levels of Treatment and conc

aggregate(uptake ~ Treatment + conc, data = CO2, FUN = summary)

##     Treatment conc uptake.Min. uptake.1st Qu. uptake.Median uptake.Mean
## 1  nonchilled   95    10.60000       11.47500      12.80000    13.28333
## 2     chilled   95     7.70000        9.60000      10.55000    11.23333
## 3  nonchilled  175    19.20000       20.05000      24.65000    25.11667
## 4     chilled  175    11.40000       15.67500      19.50000    19.45000
## 5  nonchilled  250    25.80000       27.30000      32.70000    32.46667
## 6     chilled  250    12.30000       17.95000      24.20000    25.28333
## 7  nonchilled  350    27.90000       30.45000      34.50000    35.13333
## 8     chilled  350    13.00000       18.15000      26.45000    26.20000
## 9  nonchilled  500    28.50000       31.27500      33.85000    35.10000
## 10    chilled  500    12.50000       18.30000      26.00000    26.65000
## 11 nonchilled  675    28.10000       31.42500      35.80000    36.01667
## 12    chilled  675    13.70000       19.72500      28.80000    27.88333
## 13 nonchilled 1000    27.80000       32.50000      37.60000    37.38333
## 14    chilled 1000    14.40000       20.40000      30.30000    29.78333
##    uptake.3rd Qu. uptake.Max.
## 1        15.40000    16.20000
## 2        13.30000    15.10000
## 3        29.62500    32.40000
## 4        23.32500    27.30000
## 5        36.52500    40.30000
## 6        33.82500    38.10000
## 7        40.65000    42.10000
## 8        34.45000    38.80000
## 9        39.27500    42.90000
## 10       37.07500    38.90000
## 11       40.85000    43.90000
## 12       36.97500    39.60000
## 13       43.15000    45.50000
## 14       40.72500    42.40000

• Understand types of data and their distributions

• Numerical summaries

  • Contingency Tables: table
    
  • Mean/Median: mean, median
  • Standard Deviation/Variance/IQR: sd, var, IQR
  • Quantiles/Percentiles: quantile for more general quantiles

• Across subgroups with dplyr::group_by and dplyr::summarize or aggregate

• Go to the course files page and try Exercise 9 - Numeric Summaries

• Understand types of data and their distributions

• Numerical summaries (across subgroups)

  • Contingency Tables
    
  • Mean/Median
    
  • Standard Deviation/Variance/IQR
  • Quantiles/Percentiles

• Graphical summaries (across subgroups)

  • Bar plots (categorical data)
    
  • Histograms
    
  • Box plots
    
  • Scatter plots

Three major systems for plotting:

• Base R (built-in functions)

• Lattice

• ggplot2 (sort of part of the tidyverse - Cheat Sheet)

  • ggplot(data = data_frame) creates a plot instance
    
  • Add “layers” to the system (geoms or stats)

Great reference book here!

ggplot2 basics (Cheat Sheet)

• ggplot(data = data_frame) creates a plot instance
  
• Add “layers” to the system (geoms or stats)
  • Creates a visualization of the data

ggplot2 basics (Cheat Sheet)

• ggplot(data = data_frame) creates a plot instance
  
• Add “layers” to the system (geoms or stats)
  • Creates a visualization of the data
    
• Modify layer “mapping” args (aes)
  • Ex: size, color, and x, y location(s)
    
• Coordinate system (mostly use Cartesian plane)
  
• Optional: Titles, etc.

• factor - special class of vector with a levels attribute

• Levels define all possible values for that variable
  
  • Great for variable like Day (Monday, Tuesday, …)
  • Not great for variable like Name where new values may come up

• Quite useful with plotting
  • Allows for easy labeling of subgroups

• Consider data on titanic passengers in titanic.csv

#convert survival status to a factor
titanicData$survived <- as.factor(titanicData$survived)
levels(titanicData$survived) #R knows it isn't numeric now

## [1] "0" "1"

• Can’t add value unless it is a level

titanicData$survived[1] <- "5"

## Warning in `[<-.factor`(`*tmp*`, 1, value = structure(c(NA, 2L, 1L, 1L, :
## invalid factor level, NA generated

• Useful if you want to create better labels (or change the ordering)

levels(titanicData$survived) <- c("Died", "Survived")
levels(titanicData$survived)

## [1] "Died"     "Survived"

Categorical variable - entries are a label or attribute

Generally, describe distriubtion using a barplot!

• Barplots via ggplot + geom_bar

titanicData <- read_csv("../datasets/titanic.csv")
titanicData$mySurvived <- as.factor(titanicData$survived)
levels(titanicData$mySurvived) <- c("Died", "Survived")
titanicData$myEmbarked <- as.factor(titanicData$embarked)
levels(titanicData$myEmbarked) <- c("Cherbourg", "Queenstown", "Southampton")
titanicData <- titanicData %>% drop_na(mySurvived, sex, myEmbarked)

• Barplots via ggplot + geom_bar

• Across x-axis we want our categories - specify with aes(x = ...)

ggplot(data = titanicData, aes(x = mySurvived))

• Barplots via ggplot + geom_bar

• Must add geom (or stat) layer!

ggplot(data = titanicData, aes(x = mySurvived)) + geom_bar()

• Generally: Save base object, then “add layers”

g <- ggplot(data = titanicData, aes(x = mySurvived))
g + geom_bar()

• Generally: Save base object, then “add layers”

g <- ggplot(data = titanicData, aes(x = mySurvived))
g + geom_bar()

• aes() defines visual properties of objects in the plot

        x = , y = , size = , shape = , color = , alpha = , ...

• Cheat Sheet gives most common properties for a given geom

• Generally: Save base object, then “add layers”

g <- ggplot(data = titanicData, aes(x = mySurvived))
g + geom_bar()

• aes() defines visual properties of objects in the plot

        x = , y = , size = , shape = , color = , alpha = , ...

• Cheat Sheet gives most common properties for a given geom

        d + geom_bar()

        x, alpha, color, fill, linetype, size, weight

data and aes can be set in two ways;

• ‘globally’ (for all layers) via the ggplot statement

• ‘locally’ (for just that layer) via the geom, stat, etc. layer

data and aes can be set in two ways;

• ‘globally’ (for all layers) via the ggplot statement

• ‘locally’ (for just that layer) via the geom, stat, etc. layer

#global 
ggplot(data = titanicData, aes(x = mySurvived)) + geom_bar()
#local 
ggplot() + geom_bar(data = titanicData, aes(x = mySurvived))

data and aes can be set in two ways;

• ‘globally’ (for all layers) via the ggplot statement

• ‘locally’ (for just that layer) via the geom, stat, etc. layer

#global 
ggplot(data = titanicData, aes(x = mySurvived)) + geom_bar()
#local 
ggplot() + geom_bar(data = titanicData, aes(x = mySurvived))

• To set an attribute that doesn’t depend on the data (i.e. color = 'blue'), generally place these outside of the aes

• Add better labels and a title (new layers, see cheat sheet!)

ggplot(data = titanicData, aes(x = mySurvived)) + 
  geom_bar() + 
  labs(x = "Survival Status", title = "Bar Plot of Survival for Titanic Passengers")

• Stacked barplot created by via fill aesthetic and same process

  • Create base object
    
  • Add geoms
    
  • Use aes to specify aspects of the plot

• Stacked barplot created by via fill aesthetic

• Automatic assignment of colors, creation of legends, etc. for aes elements (except with group)

ggplot(data = titanicData, aes(x = mySurvived, fill = sex)) + geom_bar()

• Add custom labels by adding more layers

ggplot(data = titanicData, aes(x = mySurvived, fill = sex)) + 
  geom_bar() +
  labs(x = "Survival Status", 
       title = "Bar Plot of Survival Status for Titanic Passengers") +
  scale_fill_discrete(name = "Sex", labels = c("Female", "Male"))

• Adjusting appropriate labeling via scale_*_discrete

aes(x = survived, fill = sex)

scale_x_discrete(labels = c("Person Died", "Person Survived"))

scale_fill_discrete(name = "Sex", labels = c("Female","Male"))

• Easy to rotate a plot with coord_flip

ggplot(data = titanicData, aes(x = mySurvived, fill = sex)) + geom_bar() +
  labs(x = "Survival Status", 
       title = "Bar Plot of Survival Status for Titanic Passengers") + 
  scale_x_discrete(labels = c("Person Died", "Person Survived")) + 
  scale_fill_discrete(name = "Sex", labels = c("Female", "Male")) + 
  coord_flip()

Note: Most geoms have a corresponding stat that can be used

geom_bar(mapping = NULL, data = NULL, stat = "count", position = "stack", ..., width = NULL, binwidth = NULL, na.rm = FALSE, show.legend = NA, inherit.aes = TRUE)

• Equivalent plots via:

ggplot(data = titanicData, aes(x = survived, fill = sex)) + geom_bar()
ggplot(data = titanicData, aes(x = survived, fill = sex)) + stat_count()

• Can modify the stat: if you have summary data, use identity

sumData <- titanicData %>% 
  group_by(survived, sex) %>% 
  summarize(count = n())
ggplot(sumData, aes(x = survived, y = count, fill = sex)) + 
  geom_bar(stat = "identity")

• Side-by-side barplot created by via position aesthetic

  • dodge for side-by-side bar plot
    
  • jitter for continuous data with many points at same values
    
  • fill stacks bars and standardises each stack to have constant height
    
  • stack stacks bars on top of each other

• Side-by-side barplot created by via position aesthetic

ggplot(data = titanicData, aes(x = mySurvived, fill = sex)) + 
  geom_bar(position = "dodge")

• position = fill stacks bars and standardises each stack to have constant height (especially useful with equal group sizes)

ggplot(data = titanicData, aes(x = mySurvived, fill = sex)) + 
  geom_bar(position = "fill")

How to create same plot for each myEmbarked value? Use faceting!

How to create same plot for each myEmbarked value? Use faceting!

facet_wrap(~ var) - creates a plot for each setting of var

• Can specify nrow and ncol or let R figure it out

How to create same plot for each myEmbarked value? Use faceting!

facet_wrap(~ var) - creates a plot for each setting of var

• Can specify nrow and ncol or let R figure it out

facet_grid(var1 ~ var2) - creats a plot for each combination of var1 and var2

• var1 values across rows

• var2 values across columns

• Use . ~ var2 or var1 ~ . to have only one row or column

How to create same plot for each myEmbarked value? Use faceting!

• facet_wrap(~ var) - creates a plot for each setting of var

ggplot(data = titanicData, aes(x = mySurvived)) + 
  geom_bar(aes(fill =sex), position = "dodge") +
  facet_wrap(~ myEmbarked)

General ggplot things:

• Can set local or global aes

• Modify titles/labels by adding more layers

• Faceting (multiple plots) via facet_grid or facet_wrap

• Only need aes if setting a mapping value that is dependent on the data (or you want to create a custom legend!)

• Go to the course files page and try Exercise 10 - Bar Plots

Numeric variables - generally, describe distribution via a histogram or boxplot!

Same process:

• Create base object
  
• Add geoms
  
• Use aes to specify aspects of the plot

• Kernel Smoother - Smoothed version of a histogram

• Common aes values (from cheat sheet):

        c + geom_density(kernel = "gaussian")

        x, y, alpha, color, fill, group, linetype, size, weight

• Only x = is really needed

• Kernel Smoother - Smoothed version of a histogram

g <- ggplot(titanicData, aes(x = age))
g + geom_density()

• Kernel Smoother - Smoothed version of a histogram

• fill a useful aesthetic!

g + geom_density(adjust = 0.5, alpha = 0.5, aes(fill = mySurvived))

• Kernel Smoother - Smoothed version of a histogram

• recall position choices of dodge, jitter, fill, and stack

g + geom_density(adjust = 0.5, alpha = 0.5, position = "stack", aes(fill = mySurvived))

• Boxplot - Provides the five number summary in a graph

• Common aes values (from cheat sheet):

        f + geom_boxplot()

        x, y, lower, middle, upper, ymax, ymin, alpha, color, fill, group, linetype, shape, size, weight

• Only x =, y = are really needed

g <- ggplot(titanicData, aes(x = mySurvived, y = age))
g + geom_boxplot(fill = "grey")

• Can add data points (jittered) to see shape of data better (or use violin plot)

g + 
  geom_boxplot(fill = "grey") +
  geom_jitter(width = 0.1, alpha = 0.3)

• Order of layers important!

g + 
  geom_jitter(width = 0.1, alpha = 0.3) + 
  geom_boxplot(fill = "grey")

• Can facet easily!

g + geom_boxplot(fill = "grey") +
  geom_jitter(width = 0.1, alpha = 0.3) + 
  facet_wrap(~ myEmbarked)

Two numerical variables

• Scatter Plot - graphs points corresponding to each observation

• Common aes values (from cheat sheet):

        e + geom_point()

        x, y, alpha, color, fill, shape, size, stroke

• Only x =, y = are really needed

• Scatter Plot - graphs points corresponding to each observation

g <- ggplot(titanicData, aes(x = age, y = fare))
g + geom_point()

• Add trend lines easily (linear and loess - a smoother)

g + geom_point() +  
  geom_smooth(aes(col = "loess")) +
  geom_smooth(method = lm, aes(col = "linear")) + 
  scale_colour_manual(name = 'Smoother', values =c('linear'='red', 'loess'='purple'), 
                      labels = c('Linear','GAM'), guide = 'legend')

• May want to add value of correlation to plot

• paste() or paste0() handy

paste("Hi", "What", "Is", "Going", "On", "?", sep = " ")

## [1] "Hi What Is Going On ?"

paste("Hi", "What", "Is", "Going", "On", "?", sep = ".")

## [1] "Hi.What.Is.Going.On.?"

paste0("Hi", "What", "Is", "Going", "On", "?")

## [1] "HiWhatIsGoingOn?"

correlation <- cor(titanicData$fare, titanicData$age, use = "complete.obs")
g + geom_point() +
  geom_smooth(method = lm, col = "Red") + 
  geom_text(x = 40, y = 400, size = 5, 
            label = paste0("Correlation = ", round(correlation, 2)))

• Text for points with geom_text

g + geom_text(aes(label = survived, color = mySurvived))

g + geom_point(aes(color = sex), size = 2.5) +
  facet_wrap(~ myEmbarked)

Many extension packages that do nice things!

library(GGally) #install GGally if needed
ggpairs(iris, aes(colour = Species, alpha = 0.4))

Numeric variable - entries are a numerical value where math can be performed

Most common plots:

• Histogram (geom_hist), Density (geom_density)

• Boxplot (geom_boxplot), Violin plot (geom_violin)

• Scatter plot (geom_point), Smoothers (geom_smooth)

• Jittered points (geom_jitter)

• Text on plot (geom_text)

General ggplot things:

• Can set local or global aes

• Modify titles/labels by adding more layers

• Use either stat or geom layer

• Faceting (multiple plots) via facet_grid or facet_wrap

• Only need aes if setting a mapping value that is dependent on the data (or you want to create a custom legend!)

• esquisse is a great package for exploring ggplot2!

• Go to the course files page and try Exercise 11 - ggplot