Week 3 Notes Ch 8
Lorraine Gaudio
2023-01-24
This document was composed from Dr. Snopkowski’s ANTH 504 Week 3 lecture and from Introduction to Data Science: Data analysis and prediction algorithms with R by Rafael A.Irizarry.
GGPLOT2: Graphing In R Markdown
There are 3 components of our ggplot2 function: 1. Data 2. Geometry – type of chart (scatterplot , barchart, histogram) 3. Aesthetic (aes) mapping – x-axis, y- axis, color
CONSTRUCTING THE MURDERS PLOT
library(dslabs)
library(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.4.0 ✔ purrr 1.0.1
## ✔ tibble 3.1.8 ✔ dplyr 1.0.10
## ✔ tidyr 1.2.1 ✔ stringr 1.5.0
## ✔ readr 2.1.3 ✔ forcats 0.5.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()data(murders)#ggplot(data=murders)
# OR
murders %>% ggplot()
Class
We can associate a plot object like this:
# p <- ggplot(data=murders)
# OR
# p <- ggplot(murders)
# OR
p <- murders %>% ggplot()What is the class of p?
class(p)## [1] "gg" "ggplot"What happens when you print p?
p
Geometries
In ggplot2, we create graphs by adding layers:
Data %>% ggplot() + LAYER1 + LAYER2 + LAYER3 + …+ LAYER n
Usually, first layer defines the geometry. In this case we want to make
a scatter plot. Let’s look at our Data Visualization (cheat) sheet and
try to find the code that represents a scatter
plot.geom_point()
Aesthetics
For geom_point() to work, we need to tell R which data we want to plot. If we look at the help file for geom_point(), and look at the “Aesthetics” section, we’ll see the arguments for this function
murders %>% ggplot() +
geom_point(aes(population, total))
murders## state abb region population total
## 1 Alabama AL South 4779736 135
## 2 Alaska AK West 710231 19
## 3 Arizona AZ West 6392017 232
## 4 Arkansas AR South 2915918 93
## 5 California CA West 37253956 1257
## 6 Colorado CO West 5029196 65
## 7 Connecticut CT Northeast 3574097 97
## 8 Delaware DE South 897934 38
## 9 District of Columbia DC South 601723 99
## 10 Florida FL South 19687653 669
## 11 Georgia GA South 9920000 376
## 12 Hawaii HI West 1360301 7
## 13 Idaho ID West 1567582 12
## 14 Illinois IL North Central 12830632 364
## 15 Indiana IN North Central 6483802 142
## 16 Iowa IA North Central 3046355 21
## 17 Kansas KS North Central 2853118 63
## 18 Kentucky KY South 4339367 116
## 19 Louisiana LA South 4533372 351
## 20 Maine ME Northeast 1328361 11
## 21 Maryland MD South 5773552 293
## 22 Massachusetts MA Northeast 6547629 118
## 23 Michigan MI North Central 9883640 413
## 24 Minnesota MN North Central 5303925 53
## 25 Mississippi MS South 2967297 120
## 26 Missouri MO North Central 5988927 321
## 27 Montana MT West 989415 12
## 28 Nebraska NE North Central 1826341 32
## 29 Nevada NV West 2700551 84
## 30 New Hampshire NH Northeast 1316470 5
## 31 New Jersey NJ Northeast 8791894 246
## 32 New Mexico NM West 2059179 67
## 33 New York NY Northeast 19378102 517
## 34 North Carolina NC South 9535483 286
## 35 North Dakota ND North Central 672591 4
## 36 Ohio OH North Central 11536504 310
## 37 Oklahoma OK South 3751351 111
## 38 Oregon OR West 3831074 36
## 39 Pennsylvania PA Northeast 12702379 457
## 40 Rhode Island RI Northeast 1052567 16
## 41 South Carolina SC South 4625364 207
## 42 South Dakota SD North Central 814180 8
## 43 Tennessee TN South 6346105 219
## 44 Texas TX South 25145561 805
## 45 Utah UT West 2763885 22
## 46 Vermont VT Northeast 625741 2
## 47 Virginia VA South 8001024 250
## 48 Washington WA West 6724540 93
## 49 West Virginia WV South 1852994 27
## 50 Wisconsin WI North Central 5686986 97
## 51 Wyoming WY West 563626 5murders %>% ggplot() +
geom_point(aes(population/10^6, total))
murders## state abb region population total
## 1 Alabama AL South 4779736 135
## 2 Alaska AK West 710231 19
## 3 Arizona AZ West 6392017 232
## 4 Arkansas AR South 2915918 93
## 5 California CA West 37253956 1257
## 6 Colorado CO West 5029196 65
## 7 Connecticut CT Northeast 3574097 97
## 8 Delaware DE South 897934 38
## 9 District of Columbia DC South 601723 99
## 10 Florida FL South 19687653 669
## 11 Georgia GA South 9920000 376
## 12 Hawaii HI West 1360301 7
## 13 Idaho ID West 1567582 12
## 14 Illinois IL North Central 12830632 364
## 15 Indiana IN North Central 6483802 142
## 16 Iowa IA North Central 3046355 21
## 17 Kansas KS North Central 2853118 63
## 18 Kentucky KY South 4339367 116
## 19 Louisiana LA South 4533372 351
## 20 Maine ME Northeast 1328361 11
## 21 Maryland MD South 5773552 293
## 22 Massachusetts MA Northeast 6547629 118
## 23 Michigan MI North Central 9883640 413
## 24 Minnesota MN North Central 5303925 53
## 25 Mississippi MS South 2967297 120
## 26 Missouri MO North Central 5988927 321
## 27 Montana MT West 989415 12
## 28 Nebraska NE North Central 1826341 32
## 29 Nevada NV West 2700551 84
## 30 New Hampshire NH Northeast 1316470 5
## 31 New Jersey NJ Northeast 8791894 246
## 32 New Mexico NM West 2059179 67
## 33 New York NY Northeast 19378102 517
## 34 North Carolina NC South 9535483 286
## 35 North Dakota ND North Central 672591 4
## 36 Ohio OH North Central 11536504 310
## 37 Oklahoma OK South 3751351 111
## 38 Oregon OR West 3831074 36
## 39 Pennsylvania PA Northeast 12702379 457
## 40 Rhode Island RI Northeast 1052567 16
## 41 South Carolina SC South 4625364 207
## 42 South Dakota SD North Central 814180 8
## 43 Tennessee TN South 6346105 219
## 44 Texas TX South 25145561 805
## 45 Utah UT West 2763885 22
## 46 Vermont VT Northeast 625741 2
## 47 Virginia VA South 8001024 250
## 48 Washington WA West 6724540 93
## 49 West Virginia WV South 1852994 27
## 50 Wisconsin WI North Central 5686986 97
## 51 Wyoming WY West 563626 5Note: aes is required Within aes you don’t need to refer to the dataset name.
Layers
We want to add a label for each point to identify the state. Looking at our cheat sheet, what function do we need to add labels (or plot text)? geom_text() with argments: x, y, label So, we can add a layer
murders %>% ggplot() + geom_point(aes(population/10^6,
total)) + geom_text(aes(population/10^6, total, label
= abb))
murders## state abb region population total
## 1 Alabama AL South 4779736 135
## 2 Alaska AK West 710231 19
## 3 Arizona AZ West 6392017 232
## 4 Arkansas AR South 2915918 93
## 5 California CA West 37253956 1257
## 6 Colorado CO West 5029196 65
## 7 Connecticut CT Northeast 3574097 97
## 8 Delaware DE South 897934 38
## 9 District of Columbia DC South 601723 99
## 10 Florida FL South 19687653 669
## 11 Georgia GA South 9920000 376
## 12 Hawaii HI West 1360301 7
## 13 Idaho ID West 1567582 12
## 14 Illinois IL North Central 12830632 364
## 15 Indiana IN North Central 6483802 142
## 16 Iowa IA North Central 3046355 21
## 17 Kansas KS North Central 2853118 63
## 18 Kentucky KY South 4339367 116
## 19 Louisiana LA South 4533372 351
## 20 Maine ME Northeast 1328361 11
## 21 Maryland MD South 5773552 293
## 22 Massachusetts MA Northeast 6547629 118
## 23 Michigan MI North Central 9883640 413
## 24 Minnesota MN North Central 5303925 53
## 25 Mississippi MS South 2967297 120
## 26 Missouri MO North Central 5988927 321
## 27 Montana MT West 989415 12
## 28 Nebraska NE North Central 1826341 32
## 29 Nevada NV West 2700551 84
## 30 New Hampshire NH Northeast 1316470 5
## 31 New Jersey NJ Northeast 8791894 246
## 32 New Mexico NM West 2059179 67
## 33 New York NY Northeast 19378102 517
## 34 North Carolina NC South 9535483 286
## 35 North Dakota ND North Central 672591 4
## 36 Ohio OH North Central 11536504 310
## 37 Oklahoma OK South 3751351 111
## 38 Oregon OR West 3831074 36
## 39 Pennsylvania PA Northeast 12702379 457
## 40 Rhode Island RI Northeast 1052567 16
## 41 South Carolina SC South 4625364 207
## 42 South Dakota SD North Central 814180 8
## 43 Tennessee TN South 6346105 219
## 44 Texas TX South 25145561 805
## 45 Utah UT West 2763885 22
## 46 Vermont VT Northeast 625741 2
## 47 Virginia VA South 8001024 250
## 48 Washington WA West 6724540 93
## 49 West Virginia WV South 1852994 27
## 50 Wisconsin WI North Central 5686986 97
## 51 Wyoming WY West 563626 5OR without a point (dots). geom_point() makes the
dot.
murders %>% ggplot() + geom_text(aes(population/10^6, total,
label = abb))
murders## state abb region population total
## 1 Alabama AL South 4779736 135
## 2 Alaska AK West 710231 19
## 3 Arizona AZ West 6392017 232
## 4 Arkansas AR South 2915918 93
## 5 California CA West 37253956 1257
## 6 Colorado CO West 5029196 65
## 7 Connecticut CT Northeast 3574097 97
## 8 Delaware DE South 897934 38
## 9 District of Columbia DC South 601723 99
## 10 Florida FL South 19687653 669
## 11 Georgia GA South 9920000 376
## 12 Hawaii HI West 1360301 7
## 13 Idaho ID West 1567582 12
## 14 Illinois IL North Central 12830632 364
## 15 Indiana IN North Central 6483802 142
## 16 Iowa IA North Central 3046355 21
## 17 Kansas KS North Central 2853118 63
## 18 Kentucky KY South 4339367 116
## 19 Louisiana LA South 4533372 351
## 20 Maine ME Northeast 1328361 11
## 21 Maryland MD South 5773552 293
## 22 Massachusetts MA Northeast 6547629 118
## 23 Michigan MI North Central 9883640 413
## 24 Minnesota MN North Central 5303925 53
## 25 Mississippi MS South 2967297 120
## 26 Missouri MO North Central 5988927 321
## 27 Montana MT West 989415 12
## 28 Nebraska NE North Central 1826341 32
## 29 Nevada NV West 2700551 84
## 30 New Hampshire NH Northeast 1316470 5
## 31 New Jersey NJ Northeast 8791894 246
## 32 New Mexico NM West 2059179 67
## 33 New York NY Northeast 19378102 517
## 34 North Carolina NC South 9535483 286
## 35 North Dakota ND North Central 672591 4
## 36 Ohio OH North Central 11536504 310
## 37 Oklahoma OK South 3751351 111
## 38 Oregon OR West 3831074 36
## 39 Pennsylvania PA Northeast 12702379 457
## 40 Rhode Island RI Northeast 1052567 16
## 41 South Carolina SC South 4625364 207
## 42 South Dakota SD North Central 814180 8
## 43 Tennessee TN South 6346105 219
## 44 Texas TX South 25145561 805
## 45 Utah UT West 2763885 22
## 46 Vermont VT Northeast 625741 2
## 47 Virginia VA South 8001024 250
## 48 Washington WA West 6724540 93
## 49 West Virginia WV South 1852994 27
## 50 Wisconsin WI North Central 5686986 97
## 51 Wyoming WY West 563626 5We might want to have dots and move the label.
TINKERING WITH ARGUMENTS
Each geometry function has many arguments that you can adjust. For example, in the geom_point() function, you can adjust the size of the points.
murders %>% ggplot() + geom_point(aes(population/10^6, total),
size =4) + geom_text(aes(population/10^6, total, label = abb))
murders## state abb region population total
## 1 Alabama AL South 4779736 135
## 2 Alaska AK West 710231 19
## 3 Arizona AZ West 6392017 232
## 4 Arkansas AR South 2915918 93
## 5 California CA West 37253956 1257
## 6 Colorado CO West 5029196 65
## 7 Connecticut CT Northeast 3574097 97
## 8 Delaware DE South 897934 38
## 9 District of Columbia DC South 601723 99
## 10 Florida FL South 19687653 669
## 11 Georgia GA South 9920000 376
## 12 Hawaii HI West 1360301 7
## 13 Idaho ID West 1567582 12
## 14 Illinois IL North Central 12830632 364
## 15 Indiana IN North Central 6483802 142
## 16 Iowa IA North Central 3046355 21
## 17 Kansas KS North Central 2853118 63
## 18 Kentucky KY South 4339367 116
## 19 Louisiana LA South 4533372 351
## 20 Maine ME Northeast 1328361 11
## 21 Maryland MD South 5773552 293
## 22 Massachusetts MA Northeast 6547629 118
## 23 Michigan MI North Central 9883640 413
## 24 Minnesota MN North Central 5303925 53
## 25 Mississippi MS South 2967297 120
## 26 Missouri MO North Central 5988927 321
## 27 Montana MT West 989415 12
## 28 Nebraska NE North Central 1826341 32
## 29 Nevada NV West 2700551 84
## 30 New Hampshire NH Northeast 1316470 5
## 31 New Jersey NJ Northeast 8791894 246
## 32 New Mexico NM West 2059179 67
## 33 New York NY Northeast 19378102 517
## 34 North Carolina NC South 9535483 286
## 35 North Dakota ND North Central 672591 4
## 36 Ohio OH North Central 11536504 310
## 37 Oklahoma OK South 3751351 111
## 38 Oregon OR West 3831074 36
## 39 Pennsylvania PA Northeast 12702379 457
## 40 Rhode Island RI Northeast 1052567 16
## 41 South Carolina SC South 4625364 207
## 42 South Dakota SD North Central 814180 8
## 43 Tennessee TN South 6346105 219
## 44 Texas TX South 25145561 805
## 45 Utah UT West 2763885 22
## 46 Vermont VT Northeast 625741 2
## 47 Virginia VA South 8001024 250
## 48 Washington WA West 6724540 93
## 49 West Virginia WV South 1852994 27
## 50 Wisconsin WI North Central 5686986 97
## 51 Wyoming WY West 563626 5The nudge_x argument moves the text slightly to the right or left
murders %>% ggplot() + geom_point(aes(population/10^6, total),
size =1) + geom_text(aes(population/10^6, total, label = abb),
nudge_x=1)
GLOBAL VS. LOCAL AESTHETIC MAPPINGS
In the previous slide, we included aes(population/10^6, total) twice. To avoid this, we can set a global aesthetic mapping so that all geometries.
murders %>% ggplot(aes(population/10^6, total, label=abb)) +
geom_point(size=1) +
geom_text(nudge_x=1.5)
murders## state abb region population total
## 1 Alabama AL South 4779736 135
## 2 Alaska AK West 710231 19
## 3 Arizona AZ West 6392017 232
## 4 Arkansas AR South 2915918 93
## 5 California CA West 37253956 1257
## 6 Colorado CO West 5029196 65
## 7 Connecticut CT Northeast 3574097 97
## 8 Delaware DE South 897934 38
## 9 District of Columbia DC South 601723 99
## 10 Florida FL South 19687653 669
## 11 Georgia GA South 9920000 376
## 12 Hawaii HI West 1360301 7
## 13 Idaho ID West 1567582 12
## 14 Illinois IL North Central 12830632 364
## 15 Indiana IN North Central 6483802 142
## 16 Iowa IA North Central 3046355 21
## 17 Kansas KS North Central 2853118 63
## 18 Kentucky KY South 4339367 116
## 19 Louisiana LA South 4533372 351
## 20 Maine ME Northeast 1328361 11
## 21 Maryland MD South 5773552 293
## 22 Massachusetts MA Northeast 6547629 118
## 23 Michigan MI North Central 9883640 413
## 24 Minnesota MN North Central 5303925 53
## 25 Mississippi MS South 2967297 120
## 26 Missouri MO North Central 5988927 321
## 27 Montana MT West 989415 12
## 28 Nebraska NE North Central 1826341 32
## 29 Nevada NV West 2700551 84
## 30 New Hampshire NH Northeast 1316470 5
## 31 New Jersey NJ Northeast 8791894 246
## 32 New Mexico NM West 2059179 67
## 33 New York NY Northeast 19378102 517
## 34 North Carolina NC South 9535483 286
## 35 North Dakota ND North Central 672591 4
## 36 Ohio OH North Central 11536504 310
## 37 Oklahoma OK South 3751351 111
## 38 Oregon OR West 3831074 36
## 39 Pennsylvania PA Northeast 12702379 457
## 40 Rhode Island RI Northeast 1052567 16
## 41 South Carolina SC South 4625364 207
## 42 South Dakota SD North Central 814180 8
## 43 Tennessee TN South 6346105 219
## 44 Texas TX South 25145561 805
## 45 Utah UT West 2763885 22
## 46 Vermont VT Northeast 625741 2
## 47 Virginia VA South 8001024 250
## 48 Washington WA West 6724540 93
## 49 West Virginia WV South 1852994 27
## 50 Wisconsin WI North Central 5686986 97
## 51 Wyoming WY West 563626 5Label can go in global or local
Or
p <- murders %>% ggplot(aes(population/10^6, total, label = abb))
p + geom_point(size=1) +
geom_text(nudge_x = 1.5)
This allows for multiple types a graphs.
OVERRIDING GLOBAL MAPPINGS
If we want to override the global mapping by defining a new mapping within each layer. These local definitions override the global.
p <- murders %>% ggplot(aes(population/10^6, total, label = abb))
p + geom_point(size = 3) + geom_text(aes(x=10, y=800, label =
"Label at 10,800"))
Scales
Our desired scale is log-scale. This needs to be changed through a
scales layer. Take a look at your cheat sheet to see if you can see how
to adjust this. There are 2 that may be useful:
scale_x_log10() and scale_y_log10() Or
scale_x_continuous() which lets us transform a value along
with setting labels, position of the axis, specifying a secondary
axis
p <- murders %>% ggplot(aes(population/10^6, total, label = abb))
p + geom_point(size=1) + geom_text(nudge_x = 0.05) + scale_x_log10() +
scale_y_log10()
### LABELS AND TITLES Looking at our cheat sheet under “Labels” we can
see that we can add labels / titles using labs
p + geom_point(size = 1) + geom_text(nudge_x = 0.05) +
scale_x_log10() + scale_y_log10() +
labs(x = "Populations in millions (log scale)",
y = "Total number of murders (log scale)",
title = "US Gun Murders in 2010") 
Categories as Colors
We can change the color of the points using the col argument in the geom_point function Let’s store our previous layers into p
p <- murders %>% ggplot(aes(population/10^6, total, label = abb)) +
geom_text(nudge_x = 0.05) +
scale_x_log10() +
scale_y_log10() +
labs(x = "Populations in millions (log scale)",
y = "Total number of murders (log scale)",
title = "US Gun Murders in 2010")
p
Now let’s add geom_point, using a color (or col) argument
p + geom_point(size = 1, color = "blue")
Categories As Colors
What we want is a different color for each region. To do this, we can assign color = region, but since this is an aesthetic mapping, we need to use aes
p + geom_point(aes(col=region), size=2)
ggplot2 automatically will add a legend. To remove it, use:
p + geom_point(aes(col = region), size = 2, show.legend=FALSE)
### Annotation, Shapes, and Adjustments We often want to add “stuff” to
our figure other than the data. For instance, lines, shaded areas,
etc.
r <- murders %>% summarize(rate = sum(total) / sum(population)*10^6)
r ## rate
## 1 30.34555class(r) ## [1] "data.frame"Let’s add a line representing the average US murder rate (per 1,000,000 people – since that’s our x-axis scale). To compute this value:
r <- murders %>%
summarize(rate = sum(total) / sum(population)*10^6) %>%
pull(rate)
class(r)## [1] "numeric"The line of the rate is defined as: y=rx, but since we are on log scale, we get: log(y) = log(r) + log(x), so our line has an intercept of log(r) and a slope of 1. To add a line, we use geom_abline (ab represents the a(intercept) and b(slope). The default is intercept = 0 and slope = 1, so in our case, we only have to define the intercept.
murders %>% ggplot(aes(population/10^6, total, label = abb)) +
geom_text(nudge_x = 0.05) +
scale_x_log10() +
scale_y_log10() +
labs(x = "Populations in millions (log scale)",
y = "Total number of murders (log scale)",
title = "US Gun Murders in 2010") +
geom_point(aes(col=region), size=2) +
geom_abline(intercept = log10(r))
lty = linetype; 0=blank, 1=solid (default), 2=dashed, 3=dotted,
4=dotdash, 5=longdash, 6=twodash
We can change the line type to a color, and we can draw it first so it doesn’t go over the points.
p <- murders %>% ggplot(aes(population/10^6, total, label = abb)) +
geom_point(aes(col=region), size=2) +
geom_text(nudge_x = 0.05) +
scale_x_log10() +
scale_y_log10() +
labs(x = "Populations in millions (log scale)",
y = "Total number of murders (log scale)",
title = "US Gun Murders in 2010") +
geom_abline(intercept = log10(r), lty=2, color ="darkgrey")
p 
Add-On Packages
There are add-on packages that further augment the ggplot2 package. We will use: ggthemes #adds a variety of themes ggrepel #positions labels
#install.packages("ggthemes")
library(ggthemes)p + theme_economist()
#install.packages("ggrepel")
library(ggrepel)p <- murders %>% ggplot(aes(population/10^6, total, label = abb)) +
geom_point(aes(col=region), size=2) +
geom_text_repel() +
scale_x_log10() +
scale_y_log10() +
labs(x = "Populations in millions (log scale)",
y = "Total number of murders (log scale)",
title = "US Gun Murders in 2010") +
geom_abline(intercept = log10(r), lty=2, color ="darkgrey") +
theme_economist()
p
Change geom_text with geom_text_repel which moves the labels so that don’t overlap Provides details on themes available: https://www.datanovia.com/en/blog/gg plot-themes-gallery/#use-ggthemes
Putting It All Together
murders %>% ggplot(aes(population/10^6, total, label= abb)) +
geom_point(aes(col=region), size=3) +
scale_x_log10() +
scale_y_log10() +
labs(x = "Populations in millions (log scale)",
y = "Total number of murders (log scale)",
title = "US Gun Murders in 2010",
col="Region") +
geom_abline(intercept=log10(r), lty=2, color="darkgrey") +
geom_text_repel() +
theme_economist()
Quick Plots
While ggplot is great for creating beautiful figures, sometimes we want to make quick plots. GGplot also has a qplot() function
qplot(log10(murders$population), log10(murders$total))## Warning: `qplot()` was deprecated in ggplot2 3.4.0.
Grids of Plots
#install.packages('gridExtra')
library(gridExtra)##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combine#dplyer::combine
p1 <- murders %>%
mutate(rate = total/population*10^5) %>%
filter(population < 2*10^6) %>%
ggplot(aes(population/10^6,
rate, label = abb)) +
geom_text() +
ggtitle("Small States")
p1
p2 <- murders %>%
mutate(rate = total/population*10^5) %>%
filter(population >= 2*10^6) %>%
ggplot(aes(population/10^6,
rate, label = abb)) +
geom_text() +
ggtitle("Large States")
p2
Place the graphs side by side with ncol().
grid.arrange(p1, p2, ncol = 2)
Defalt is in rows.
grid.arrange(p1, p2)