04_viz.ipynb

Kernel: R

DSCI 100 - Introduction to Data Science

Lecture 4 - Data visualization in R

2019-01-23

Housekeeping

• Grades are coming! Thanks for your patience!!!

• Quiz next week!

  • 45 min

  • open book (but not collaborative!)

  • in class, but on Canvas

  • you will get some practice quiz questions by the end of the week

Reminder

Where are we? Where are we going?

image source: R for Data Science by Grolemund & Wickham

The basic ggplot call:

plot_object <- ggplot(dataframe, aes(x = a_column, y = another_column)) + 
    geom_something()
    
plot_object

...

plot_object <- plot_object + 
    ...

Where to get help (and ideas) for creating ggplot2 visualizations?

• https://www.r-graph-gallery.com/portfolio/ggplot2-package/

Only make the plot area as big as needed!

• the default size is ridiculous!

library(tidyverse)
too_big <- ggplot(mtcars, aes(x = hp, y = mpg)) + 
    geom_point() +
    xlab("Horsepower") +
    ylab("Miles per gallon")

── Attaching packages ─────────────────────────────────────── tidyverse 1.2.1 ──
✔ ggplot2 3.1.0     ✔ purrr   0.2.5
✔ tibble  1.4.2     ✔ dplyr   0.7.7
✔ tidyr   0.8.0     ✔ stringr 1.3.1
✔ readr   1.1.1     ✔ forcats 0.3.0
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()

too_big

MIME type unknown not supported

• use the repr package to set your plot size with R in Jupyter

library(repr)
options(repr.plot.width = 2.5, repr.plot.height = 2.5)

too_big

MIME type unknown not supported

Don’t adjust the axes to zoom in on small differences (if the difference is small, show that its small!)

not_a_big_deal <- ggplot(iris, aes(x = Species, y = Sepal.Length)) + 
    geom_boxplot() +
    ylab("Sepal length")

not_a_big_deal

MIME type unknown not supported

not_a_big_deal <- not_a_big_deal +
    ylim(c(0, 7))

not_a_big_deal

Warning message:
“Removed 12 rows containing non-finite values (stat_boxplot).”

MIME type unknown not supported

Show the data (don’t hide the shape/distribution of the data behind a bar)

next two slides borrowed from Jeff Leek

Be wary of overplotting...

too_much <- ggplot(diamonds, aes(x = carat, y = price)) +
    geom_point() +
    xlab("Price (US dollars)") +
    ylab("Size (carat)")

too_much

MIME type unknown not supported

an_improvement <- ggplot(diamonds, aes(x = carat, y = price)) +
    geom_point(alpha = 0.01) +
    ylab("Price (US dollars)") +
    xlab("Size (carat)")

an_improvement

MIME type unknown not supported

Use colors sparingly

Use legends and labels so that your visualization is understandable without reading the surrounding text

Ensure the text on your visualization is big enough to be easily read

Do not use pie charts!

Do not use 3D!

Attribution

• images without code mostly come from Fundamentals of Data Visualization by Claus O. Wilke

Go and create!

Make an effective plot!

Can petal length and petal width be used to separate the Iris flower species? Create a plot to answer this question!

head(iris)

# solution
library(tidyverse)
options(repr.plot.width = 5, repr.plot.height = 3)
plot <- ggplot(iris,aes(x = Petal.Length, y = Petal.Width)) + 
    geom_point(aes(color = Species, shape = Species)) +
    xlab("Petal Length") +
    ylab("Petal Width")
plot

MIME type unknown not supported

What did we learn today

• Combine color and shape for points to separate groups in scatter plots

• Rarely use 3D or pie charts, there are other, more effective ways of communicating the information

• Some rules of thumb/guidelines for making visualizations

• How to negatively filter using !

• How to use options from the repr packages to set plot size (and that we should set plot size)!