|
Dee Chiluiza, PhD Northeastern University Boston, Massachusetts |
| Short manual series: Strategies to present graphs and images |
Purpose:
To present and combine multiple graphs and tables for better presentation.
List of topics
Insert images
- Control size and insert images using knitr::include_graphics()
- How to reduce white space around figures
- Using grid.extra::grid.arrange()
- Using patchwork library
- Use patchwork to control width of each figure
- Add annotations using Geom Text and ggrepel
- Combination of fig.chunk and HTML codes.
- Highlight data in graphs using geom_mark
- Using Colors: RGB, RColorBrewer, Terrain colors, Custom colors.
Control size and insert images using knitr::include_graphics()
# {r, fig.align = 'center', out.width="40%", out.height="50%", fig.width = 8, fig.height = 6}
knitr::include_graphics("Images/bruno.JPG") 
# Data sets (add importing codes here)
library(readxl)
library(readr)
M2Data <- read_excel("DataSets/M2Project_V2.xlsx")
t4Africa = dplyr::filter(M2Data, Market=="Africa")
t4table = table(t4Africa$Product_Category)Figure chunk. Compare size of left and right figures.
# {r, dev='png', fig.align='center', out.width="100%"}
par(mfrow=c(1,2))
t4pie1 = pie(table(t4Africa$Product_Category))
par(mar = c(0.1, 0.1, 0.1, 0.1)) # Reduce space around plots
t4pie2 = pie(table(t4Africa$Product_Category))
Using grid.extra::grid.arrange()
Check the grid.arrange() code after graphs were created. Also notice that graphs require names.
aa1 = ggplot(data=faithful,
mapping = aes(x=eruptions, y=waiting))+
geom_point()
# Different grammar
aa2= ggplot(faithful)+
geom_point(aes(x=eruptions, y=waiting))
# Mapping color
aa3= ggplot(faithful)+
geom_point(aes(x=eruptions,
y=waiting,
color=eruptions<3))
# Set color, not mapping, outside aes()
aa4=ggplot(faithful)+
geom_point(aes(x=eruptions,
y=waiting),
color="red")
grid.arrange(aa1, aa2, aa3, aa4, ncol=2, nrow=2)
aa3= ggplot(faithful)+
geom_point(aes(x=eruptions,
y=waiting,
color=eruptions<3))
aa4=ggplot(faithful)+
geom_point(aes(x=eruptions,
y=waiting),
color="red")
grid.arrange(aa3, aa4, nrow=1)
One over two layout
mpg_counted = mpg %>%
count(class, name = 'count')
barr1 = ggplot(mpg_counted) +
geom_bar(aes(x = class, y = count), stat = 'identity')
# It's the same as the code below
barr2 = ggplot(mpg) +
geom_bar(aes(x = class))
# Flip and remove background
barr3 = ggplot(mpg) +
geom_bar(aes(x = class)) +
coord_flip() +
theme_minimal()+
labs(x = "Counting", y = "Car Classes")
grid.arrange(arrangeGrob(barr1, barr2, ncol=2),
barr3, ncol=1)
Using patchwork library
library(patchwork)
p1 <- ggplot(msleep) +
geom_boxplot(aes(x = sleep_total, y = vore, fill = vore))
p2 <- ggplot(msleep) +
geom_bar(aes(y = vore, fill = vore))
p3 <- ggplot(msleep) +
geom_point(aes(x = bodywt, y = sleep_total, colour = vore)) +
scale_x_log10()
p1 + p2 + p3
The code below can be p1/ (p2|p3) or p1/ (p2+p3).
Another way.
p_all <- p_all & theme(legend.position = 'none')
p_all + plot_annotation(
title = 'Mammalian sleep patterns',
tag_levels = 'A'
)
Use patchwork to control width of each figure
Change the width of each individual figure.
Add annotations using Geom Text and ggrepel
Standard geom_text will often result in overlapping labels.
ggrepel takes care of that, link: ggreprl.
- geom_text_repel()
- geom_label_repel()
Link to ggrepel examples: https://ggrepel.slowkow.com/articles/examples.html.
library(ggrepel) # From package ("ggrepel")
ggplot(mtcars, aes(x = disp, y = mpg)) +
geom_point(color = 'red') +
geom_text_repel(aes(label = row.names(mtcars)), size=3)+
theme_classic(base_size = 16)
Combination of fig.chunk and HTML codes.
# Data sets (add importing codes here)
library(readxl)
library(readr)
M2Data <- read_excel("DataSets/M2Project_V2.xlsx")
t4Africa = dplyr::filter(M2Data, Market=="Africa")
t4table = table(t4Africa$Product_Category)
Figure chunk.
par(mar = c(0.1, 0.1, 0.1, 0.1)) # Reduce space around plots
t4pie = pie(table(t4Africa$Product_Category))
|
|
Highlight data in graphs using geom_mark
To highlight certain parts of the data, and to add annotations, use the geom_mark_*() family of geoms.
library(ggforce)
ggplot(mtcars, aes(x = disp, y = mpg)) +
geom_point(color="#A11515") +
geom_mark_ellipse(aes(filter = gear == 4,
label = '4 gear cars',
description = 'Cars with fewer gears
tend to both have higher yield and
lower displacement'))+
theme_classic(base_size = 16)
Using Colors
RGB Color Codes Chart: https://www.rapidtables.com/web/color/RGB_Color.html
# {r, fig.align = 'center', out.width="50%", out.height="20%"}
knitr::include_graphics("Images/rgb.JPG") 
How to use colors on graphs.
- I have three favorite color palettes. one is RColorBrewer.
To see all colors use display.brewer.all(), this will display an image with all color sets and their corresponding names.
Using this palete, you can select the color set ant the number of color you need.
For example: If we have a graph with 6 bars, we can have a preview of the color combination by calling the color set and number of colors needed. display.brewer.pal(6,“Dark2”)
# {r col1, fig.align = 'center', out.width="60%", fig.asp = 1}
par(mai=c(0.2,0.6,0.2,0.2), mfcol=c(1,3), mar=c(0.2,3,0.2,0.2))
col1a = display.brewer.all()
col1b = display.brewer.pal(8,"Dark2")
col1c = display.brewer.pal(9,"Greens")
Use the RColorBrewer() in bar plots.
# {r col2, fig.align="center"}
par(mfrow=c(1,2), mar=c(1,2,1.2,1.2), mai=c(0.2,0.5,0.2,0.2))
barcol = c(10, 8, 6, 5, 7, 9)
col2a = barplot(barcol,
col = brewer.pal(6,"Dark2"),
las=1,
ylim=c(0,12))
col2b = barplot(barcol,
col = brewer.pal(6,"Greens"),
las=1,
ylim=c(0,12))
| Custom
Make your own palettes.
par(mfrow=c(1,2), mar=c(1,2,1.2,1.2), mai=c(0.2,0.4,0.6 ,0.2))
# My colors option 1
mycolors1 = c("blue", "yellow", "green", "pink", "lightblue", "cyan")
# My colors option 2
mycolors2 = c("#E96464", "#E564E9", "#A764E9", "#6483E9", "#64E9A7", "#F7F36D")
# Application
col2a = barplot(barcol,
col = mycolors1,
main= "Using mycolors 1",
las=1,
ylim=c(0,12))
col2b = barplot(barcol,
col = mycolors2,
las=1,
main= "Using my colors 2",
ylim=c(0,12))
Disclaimer
Disclaimer: This short series manual project is a work in progress. Until otherwise clearly stated, this material is considered to be draft version.
June 2021
Last update: 12 October, 2021
Boston, Massachusetts, USA
