Exploring Titanic Data

• Data: passengers
  Don't have passengers?
  Use import-passengers.R
to get it.
• Code: demo-plots.R
• Will run this code together
• ~30 Minutes
• Followed By: A Break!

Learning Objectives

• R has many graphical packages
  • ggplot2
  • ggvis
  • lattice
  • plotly
  • htmlwidgets
  • rbokeh
• Today, we use Base R
• Why isn't there a graphic on a page about graphics?

Base R Graphics

• Encourage interactive use
• Easy to remember commands
• Good defaults!
• Thank You: John W. Tukey

Opening Discussion

Plotting functions we will use today:

• plot()
  You've already used this command
• barplot()
• boxplot()
• hist()
• pairs()

These commands should be mostly self-explanatory

Age

## Remember:
mean(passengers$age)
[1] 30.31759
        

## Only one variable?
## X axis is row number.
## Can look like a trend.
plot(passengers$age)
        

## Use sample() to avoid.
plot(sample(passengers$age))
        

Age by Passenger Class

## X/Y Axis order matters!
plot(passengers$passenger_class,
     passengers$age
    )
        

Remember Factors?

• Superficially similar to character/string variables
• Tells R a variable is a categorical variable
• Factors have different defaults in plot()

Age by Passenger Class (Factor)

## Now with boxplots!
plot(as.factor(passengers$passenger_class),
     passengers$age
    )
        

Barplots

You can't create a barplot with raw data. It needs a table.

# Our passenger data has more males than females
table(passengers$sex)
female   male 
   107    143 
    

# Creates a table called tbl_sex
tbl_sex <- table(passengers$sex)
tbl_sex
female   male 
   107    143 
    

Passenger Sex

## Input is a table.
barplot(tbl_sex)
        

## Input is the output of table()
barplot(table(passengers$sex))
        

STOP! (Teachable Moment)

Your Turn!

## Learn how to use the prop.table command.
## Use this command to build a table showing the
## proportions of males and females in passengers.
?prop.table
    

prop.table(tbl_sex)*100

female   male 
  42.8   57.2 
    

Your Turn!

## Now use prop.table to build a proportional barplot.
    

barplot( prop.table(tbl_sex)*100 )
    

Hint: There's a nice plot on the next slide

Nice!

GET SOME HELP!

• Eventually, we all need some help
• Stack Overflow
• Google
• Let's practice finding help

Your Turn!

• Get familiar with this data set. We aren't done with it.
• The following two files include examples of histograms, density plots, etc. Things that might interest you.
• exploratory-plots.R
• advanced-plots.R
• Take a short break before diving into logistic regression
• All materials are accessible outside of this workshop

Take A Break!

Titanic in Cobh Harbour, County Cork Ireland