Data Visualization is an essential component of your skillset as a Data Scientist or Data Analyst. Data Visualization is basically a form of Visual communication.

ggplot2 is a plotting package that helps us to create complex plots from data in data frame.

ggplot2 functions built step by step by adding new elements

Install ggplot2 package

# install ggplot2

install.packages(ggplot2)

Load ggplot2 package

# include ggplot2 library

library(ggplot2)

During this discussion, we are going to use mtcars package for the dataset.
Note:
The matcars dataset contains information about 32 cars from 1973 motor trends magazine. The dataset is small but contains a variety of continuous and categorical variables.

Before describing ggplot2 in more detail just have a look mtcars dataset using str() command.

#structure of matcarsbasically
str(mtcars);

OutPut:

Have a look ggplot2 example 

Example:

# include ggplot2 library
library(ggplot2)
ggplot(mtcars , aes(x=wt, y=mpg))+geom_point()

OutPut:

Some points regarding ggplot2ppp

• VisualizationVisual elements in ggplot2 are called geoms (as in geometric objects bars, points …)
• The appearance and location of these geoms (size, color) are controlled by aesthetic properties.basicallybasically
• aesthetic properties are shown by aes()
• Variable that you want to plot is represented by aes() as shown in the previous example.

Goem layer	Description
geom_bar()	Create a layer with bars representing different statistical properties.
geom_point()	Create a layer with data points.
geom_line()	Create a layer with a straight line.
geom_smooth()	Create a layer with smoother.
geom_histogram()	Create a layer with a histogram.
geom_blogplot()	Create a layer with text in it.
geom_text()	Create a layer with a text in it.
geom_error_bar()	Create a layer with error bars in it.
geom_hline and geom_vline()	Create a layer with a user-defined horizontal and vertical line respectively.

How to derive iris.tidy from iris?

library(tidyr)
#Convert iris to iris.tidy using tidy function
iris.tidy <- iris %>%
  gather(key, Value, -Species) %>%
  separate(key, c("Part", "Measure"), "\\.")

print(head(iris.tidy))

How to derive iris.wide from iris?

# Load the tidyr package
library(tidyr)
# Add column with unique ids (don't need to change)
iris$Flower <- 1:nrow(iris)
# Produce the iris.wide dataset
iris.wide <- iris %>%
  gather(key, value, -Species, -Flower) %>%
  separate(key, c("Part", "Measure"), "\\.") %>%
  spread(Measure, value)

OutPut: