DATA VISUALIZATION WITH GGPLOT2. Bar Plots

Size: px

Start display at page:

Download "DATA VISUALIZATION WITH GGPLOT2. Bar Plots"

Joan White
6 years ago
Views:

1 DATA VISUALIZATION WITH GGPLOT2 Bar Plots

2 Chapter Content Common pitfalls Best way to represent data

3 Bar plot Two types Absolute values Distribution

4 Mammalian sleep > str(sleep) 'data.frame': 76 obs. of 3 variables: $ vore : Factor w/ 4 levels "Carnivore","Herbivore",..: $ total: num $ rem : num NA NA.6...

5 Dynamite plot > d <- ggplot(sleep, aes(vore, total)) + scale_y_continuous("total sleep time (h)", limits = c(, 24), breaks = seq(, 24, 3), expand = c(, )) + scale_x_discrete("eating habits") + theme_classic() > d + stat_summary(fun.y = mean, geom = "bar", fill = "grey5") + stat_summary(fun.data = mean_sdl, mult = 1, geom = "errorbar", width =.2) Total sleep time (h) Carnivore Herbivore Insectivore Omnivore Eating habits

6 Individual data points > d + geom_point(alpha =.6, position = position_jitter(width =.2)) Total sleep time (h) Carnivore Herbivore Insectivore Omnivore Eating habits

7 errorbar > d + geom_point(alpha =.6, position = position_jitter(width =.2)) + stat_summary(fun.y = mean, geom = "point", fill = "red") + stat_summary(fun.data = mean_sdl, mult = 1, geom = "errorbar", width =.2, col = "red") Total sleep time (h) Carnivore Herbivore Insectivore Omnivore Eating habits

8 pointrange > d + geom_point(alpha =.6, position = position_jitter(width =.2)) + stat_summary(fun.data = mean_sdl, mult = 1, width =.2, col = "red") Total sleep time (h) Carnivore Herbivore Insectivore Omnivore Eating habits

9 Without data points > d + stat_summary(fun.y = mean, geom = "point") + stat_summary(fun.data = mean_sdl, mult = 1, geom = "errorbar", width =.2) Total sleep time (h) Carnivore Herbivore Insectivore Omnivore Eating habits

10 Total sleep time (h) Carnivore Herbivore Insectivore Omnivore Eating habits

11 DATA VISUALIZATION WITH GGPLOT2 Let s practice!

12 DATA VISUALIZATION WITH GGPLOT2 Pie Charts

13 Stacked bar chart > ggplot(mtcars, aes(x = factor(1), fill = factor(cyl))) + geom_bar(width = 1) 3 count 2 1 factor(cyl) factor(1)

14 pie chart > ggplot(mtcars, aes(x = factor(1), fill = factor(cyl))) + geom_bar(width = 1) > ggplot(mtcars, aes(x = factor(1), fill = factor(cyl))) + geom_bar(width = 1) + coord_polar(theta = "y") 3 3 count 2 1 factor(cyl) factor(1) factor(cyl) count 1 factor(1)

15 Parts-of-a-whole 3 factor(1) 1 1 factor(cyl) count

16 HairCol > HairCol Hair Sex Value fillin n nprop 1 Black Male 56 # Brown Male 143 #A Red Male 34 #E41A1C Blond Male 46 #FFFF Black Female 52 # Brown Female 143 #A Red Female 37 #E41A1C Blond Female 81 #FFFF

17 HairCol - Bar Charts > ggplot(haircol, aes(x = Hair, y = Value, fill = fillin)) + geom_bar(stat = "identity", position = "dodge") + facet_grid(. ~ Sex) + scale_fill_identity() + theme_classic() 15 Male Female Hard to reveal interesting trends Difference in total counts is unclear 1 Value 5 Black Brown Red Blond Black Brown Red Blond Hair

18 HairCol - Pie Charts > ggplot(haircol, aes(x = n/2, y = Value, fill = fillin, width = n)) + geom_bar(stat = "identity", position = "fill") + facet_grid(. ~ Sex) + scale_fill_identity() + coord_polar(theta = "y") + theme(...) Male Female angle, area, length mediocre encoding elements

19 Alternative > ggplot(haircol, aes(x = Sex, y = Value, fill = fillin, width = nprop)) + geom_bar(stat = "identity", position= "fill") + scale_y_continuous("proportion") + scale_x_discrete("", expand = c(, )) + scale_fill_identity() + coord_flip() + theme(...) Female Male Proportion

20 DATA VISUALIZATION WITH GGPLOT2 Let s practice!

21 DATA VISUALIZATION WITH GGPLOT2 Heat Maps

22 barley.s > head(barley.s, 15) variety site Grand Rapids Duluth University Farm Morris Crookston Waseca Grand Rapids Duluth University Farm Morris Crookston Waseca Grand Rapids Duluth University Farm

23 Variety Grand Rapids Duluth University Farm Morris Crookston Waseca Year yield

24 barley > head(barley, 15) yield variety year site University Farm Waseca Morris Crookston Grand Rapids Duluth University Farm Waseca Morris Crookston Grand Rapids Duluth University Farm Waseca Morris

25 Variety Grand Rapids Duluth University Farm Morris Crookston Waseca Year yield

26 Variety Grand Rapids Duluth University Farm Morris Crookston Waseca Yield (bushels/acre) Year

27 7 Grand Rapids Duluth University Farm Morris Crookston Waseca 6 Yield (bushels/acre) Variety Year

28 7 6 5 Yield (bushels/acre) 4 3 Site Waseca Crookston Morris University Farm Duluth Grand Rapids Year

29 7 6 5 Yield (bushels/acre) 4 3 Site Waseca Crookston Morris University Farm Duluth Grand Rapids Year

30 DATA VISUALIZATION WITH GGPLOT2 Let s practice!

Analyzing Categorical Data & Displaying Quantitative Data Section 1.1 & 1.2

Analyzing Categorical Data & Displaying Quantitative Data Section 1.1 & 1.2 Reference Text: The Practice of Statistics, Fourth Edition. Starnes, Yates, Moore Starter Problem Antoinette plays a lot of golf.