R Visualization with ggplot What is ggplot Package

What is ggplot? • Package specifically for data visualization • Created by Hadley Wickham

Key Distinguishing features • ggplot 2 allows the user to add, remove or alter

The different parts to the plotting function we will cover today • Data •

Instead of using qplot, use ggplot for the two examples I showed previously

Your Turn! Create a qplot of Old Faithful’s eruption time vs. waiting time

Aesthetic Inputs • ggplot will not work unless there a set of aesthetic inputs

Aesthetics Color (Still qplot) The color input allows you to distinguish your data by

Aesthetics Color (Still qplot) For just one color, use the I() notation

Your Turn! Make a plot of old faithful’s eruptions vs waiting time with all

Aesthetics Shape (Still qplot) The shape of the individual points on the plot can

Aesthetics Size (Still qplot) Same can be done with the size of the points

Aesthetics Size (Still qplot) To make the previous slide look better, we can combine

Aesthetics Size (Still qplot) For just one size, use the I() notation

Your Turn • Spend two to three minutes playing around with different sizing, shapes,

Aesthetics for ggplot The key difference between ggplot and qplot is the x and

Aesthetics for ggplot Notice the different methods of aesthetic arguments in ggplot

Your Turn! Use ggplot to do the same old faithful plot we’ve been using

Your Turn! Now with the airquality data, use ggplot to plot ozone vs temp

Statistical Plotting (white noise) You can add ‘white noise’ to the data as well

Your Turn • Produce a jittery plot of the ozone vs temperature in the

Statistical Plotting (geom) Notice the different methods of inputting geom arguments in ggplot

Your Turn • Use ggplot to make the same plots you made previously of

Statistical Plotting (geom histogram) You can alter the characteristics of display within statistical plots

Statistical Plotting (geom density) Utilizing geom parameters with color/other aesthetic parameters can help differentiate

Statistical Plotting (adding lines for analysis)

Statistical Plotting (adding lines for analysis) This one makes more sense

Your Final Turn! • Create some nice cool-looking plots of the airquality dataset (keep

Slides: 42

Download presentation

R Visualization with ggplot

What is ggplot? • Package specifically for data visualization • Created by Hadley Wickham in 2005 (Rice professor!!!!) • Open source, most popular visualization tool in R • ggplot 2 extension

Key Distinguishing features • ggplot 2 allows the user to add, remove or alter components in a plot at a high level of abstraction • More complexity in visualization schemes available compared to base R functions

ggplot 2 examples

Plot vs ggplot

The different parts to the plotting function we will cover today • Data • Aesthetic Mappings • Geometric Objects • Statistical Transformations • Scales

Loading it into R

Analyzing the data we will use today

Basic Plots (data)

Instead of using qplot, use ggplot for the two examples I showed previously

Your Turn! Create a qplot of Old Faithful’s eruption time vs. waiting time

Aesthetic Inputs • ggplot will not work unless there a set of aesthetic inputs • color, size, and shape (visual properties that affect the display) • Other plot functions don’t have this issue because they don’t have specific inputs for display characteristics

Aesthetics Color (Still qplot) The color input allows you to distinguish your data by color gradients

Aesthetics Color (Still qplot)

Aesthetics Color (Still qplot) For just one color, use the I() notation

Your Turn! Make a plot of old faithful’s eruptions vs waiting time with all the points orange Change the color to the value of eruption time

Aesthetics Shape (Still qplot) The shape of the individual points on the plot can also be changed and utilized to reflect data characteristics

Aesthetics Size (Still qplot) Same can be done with the size of the points

Aesthetics Size (Still qplot) To make the previous slide look better, we can combine input parameters

Aesthetics Size (Still qplot) For just one size, use the I() notation

Your Turn • Spend two to three minutes playing around with different sizing, shapes, and coloring with the mpg and airquality data set • Airquality dataset is new. Spend some time to familiarize yourself with it

Aesthetics for ggplot The key difference between ggplot and qplot is the x and y coordinate values are included in specified aesthetic arguments

Aesthetics for ggplot Notice the different methods of aesthetic arguments in ggplot

Aesthetics for ggplot

Your Turn! Use ggplot to do the same old faithful plot we’ve been using (color is orange)

Your Turn! Now with the airquality data, use ggplot to plot ozone vs temp with color as the month and size as the day

Statistical Plotting (white noise) You can add ‘white noise’ to the data as well as modify the data for more accurate analysis through the geom argument

Statistical Plotting (boxplot)

Your Turn • Produce a jittery plot of the ozone vs temperature in the airquality data set • Make a boxplot

Statistical Plotting (geom) Notice the different methods of inputting geom arguments in ggplot

Statistical Plotting (geom boxplot)

Your Turn • Use ggplot to make the same plots you made previously of airquality

Statistical Plotting (geom freqpoly)

Statistical Plotting (geom histogram) You can alter the characteristics of display within statistical plots

Statistical Plotting (geom histogram)

Statistical Plotting (geom density) Utilizing geom parameters with color/other aesthetic parameters can help differentiate between data

Statistical Plotting (adding lines for analysis)

Statistical Plotting (adding lines for analysis) This one makes more sense

Labels (qplot)

Labels (ggplot)

Your Final Turn! • Create some nice cool-looking plots of the airquality dataset (keep in mind the plots must depict some type of coherent analysis) • That’s all the instruction you have. Nicest plot wins a prize!