Nilanjan Chatterjee
February, 2020
Technique to communicate insights from data through visual representation.
Allow easy understanding of large dataset.
Provides basic knowledge about variables.
Most efficient way to identify, locate, manipulate, format, and present data.
data(mtcars)
plot(mpg~wt, mtcars, pch=19, col="blue")
plot vs ggplot
Pros | Cons |
---|---|
In-built | Additional package |
Easy to learn | Steep learning curve |
Indepenedent of data-structures | Works only with data-frame |
Easy for simple plots | Verbose for complex plots |
Low level of abstraction | High abstraction level |
Visually less appealing | Visually more appealing |
Based on Grammer of graphics (Wilkinson, 2005).
Consists of several building blocks like a sentence.
#install.packages("ggplot2", dependencies = T)
library(ggplot2)
ggplot(mtcars, aes(x= wt, y= mpg))+
geom_point(colour="blue", size=3)
ggplot(mtcars) #data
ggplot(mtcars, aes(x= wt, y= mpg)) #data+aesthetic map
ggplot(mtcars, aes(x= wt, y= mpg))+ #data+aesthetic map
geom_point() #geometric obj
ggplot(mtcars, aes(x= wt, y= mpg))+ #data+aesthetic map
geom_point(colour="blue", size=3) #geometric obj
ggplot(mtcars, aes(x= wt, y= mpg))+ #data+aesthetic map
geom_point(colour="blue", size=3)+ #geometric obj
ggtitle("Scatterplot") #Plot title
ggplot(mtcars, aes(x=mpg))+
geom_bar()
ggplot(mtcars, aes(x=cyl, y=mpg, fill= cyl))+
geom_bar(stat="identity")
ggplot(mtcars, aes(x=cyl, y=mpg))+
geom_point(stat="identity", size=4)
You can export any plots using the plot window from R/RStudio.
To save files in high-resolution these commands are helpful
sct <-ggplot(mtcars, aes(x= wt, y= mpg))+
geom_point(colour="blue", size=3)+ ggtitle("Scatterplot")
ggsave(sct, "Scatterplot_with_R.jpeg", dpi=100, device = "jpeg")