Chapter-5:- Scatter Plot in R-programming

 Chapter-5:- Scatter Plot in R-programming

Example-1:- 

Create a scatter plot of car weight on x-axis and miles per gallon on y-axis by pre-program data "mtcars" in R-language

Code:-

mtcars

plot(mtcars$wt, mtcars$mpg,

     main = "Car Weight vs Fuel Efficiency",

     xlab = "car weight",

     ylab = "Miles Per Gallon",

     pch = 19,         # Solid circle shape

     col = "blue", # Change point color

     frame = FALSE)    # Remove the surrounding box

Result:-

Explanation:-

• In 1st line, "mtcars" is inbuilt program or database in R-programming language, you can in result image with yellow square box, it contain multiple car model with mpg, disp, wt, gear, carb etc. 
• In 2nd line plot() use, we discuss this already in previous chapter it contain multiple parameter like col, lwd, lty, col, xlim, ylim etc.
• In 2nd line, you see mtcars$wt, here you see $ sign, it use mainly to collect data from specific column like here we use wt, so it collect data of wt column of different model of cars
• Also we use mtcars$mpg, it means it collect data from mpg column of database
• In last line, frame if TRUE it make a box around it, if FALSE then you see no box surround the result image

Example-2:-

Plot sepal length vs sepal width for all species, set the x-axis range from 4 to 8 and y-axis from 2 to 4.5 by pre-program database "iris"

Code:-

iris

plot(iris$Sepal.Length, iris$Sepal.Width,

     col = c("blue", "yellow"),    # For color of the sepal

     pch = 19,                    # solid circle denote by 19

     main = "Custom Species Colors")   #main title

Result:-

Explanation:-

• Iris is also pre-program or inbuilt data base in R-programming language like the "mtcars"
• You can see in result image in yellow square box, "iris" contain 5 different column like Sepal.length, Sepal.width, species etc.
• If you want to collect data from specific column like Sepal.Length then we use $ sign as you can see in 2nd line code
• For use different color, use c(1st color name, 2nd color name), here c is compulsary to use, If you read previous chapter then no difficulty to understand
• pch- use for different shape/ symbol to denote
• main- for main title 

Example-3:- 

If in above example if ask to arrange the data according to petal width (means according to petal width the plot size increase or decrease, higher the petal width more size of plot in result graph)

Code:-

iris

plot(iris$Petal.Width,

     col = "red",    # For color of the sepal

     pch = 19,                    # solid circle denote by 19

     main = "Custom Species Colors",   #main title

     cex=iris$Petal.Width,    #Increase solid circle size according to value

     las=1,     # To make y-axis text horizontal

     frame.plot=FALSE)   #Remove top and right border of box

Result:-

Explanation:-

• Compare to previous example here additional command "cex" , frame.plot
• All the point already explain with comment in code

Example-4:-

Create a scatter plot of girth vs volume. Before plotting the points, add a light grey background grid using panel.first= grid() by using pre-program feature "trees", also add regression line to show the trend

Code:-

trees   #Load the database

plot(trees$Girth, trees$Volume,

     panel.first= grid(col="grey",lty="dotted"), #Create grid with background grey and dotted

     col="red",pch=20,cex=1.5)

#Add a red regression line

abline(lm(Volume ~ Girth, data = trees), col = "blue", lwd = 2)

Result:-

Explanation:-

• "trees" also inbuilt database in R-programming

Component	Function	Details
abline()	Draws a line	Function that draws a straight line on the plot
lm()	Fits a linear model	Calculate the best fit straight line that minimize the distance between line and your data points
Volume ~ Girth	Model formula	Means volume vs girth graph (use here “~”)
data	Specifies data	Like in this case trees which is inbuilt data
col	Color	Set the color of line
lwd	Line width	Sets the line width which default is 1

Example-5:- Code using ggplot2

Code:-

library(ggplot2)

# Load data

data(trees)

# Initialize plot and add styled points, Use + to recognize the next line

ggplot(trees, aes(x = Height, y = Volume)) +

geom_point(size = 4, shape = 21, fill = "green", color = "white") +

# Add orange regression line without confidence band

 geom_smooth(method = "lm", color = "orange", se = FALSE)

Result:-

Explanation:- 

• In 7th line you see "+ sign" which for recognize the next line
• geom_point to create scatter plot (point library contain color, fill, size, shape, alpha)
• geom_smooth is use to add regression line
• Further code explain below

Example-6:-

Code:-

library(ggplot2)

# Load data

data(trees)

# Initialize plot and add styled points

ggplot(trees, aes(x = Height, y = Volume)) +

geom_point(size = 4, shape = 21, fill = "darkgreen", color = "white") +

  

# Add orange regression line with confidence band by add se=TRUE

geom_smooth(method = "lm", color = "orange", se =TRUE) +

  

# Apply a clean theme and customize grid lines

  theme_minimal() +   # Sets the base line first

  theme(

    panel.grid.major = element_line(color = "lightgrey", linetype = "dashed"),

    panel.grid.minor = element_blank() # Clean up minor grid lines

  ) +

  labs(title = "Cherry Tree Height vs. Volume",

       x = "Height (ft)",

       y = "Volume (cubic ft)")

Result:-

Explanation:-

• Here, a change compare to previous example- In 11th line we use here se=TRUE instead of FALSE , so you can see gray color shadow around the orange line
• In 14th line we use theme_minimal to set the base line first
• In 15th line we use the theme command which you can see below as all code explain of theme command
• In 11th line you see "+" sign, also same you can see in 18th its mainly to recognize the next line

1. The Global Mapping (aes)

The aes() (aesthetics) function maps your data variables to visual properties.

x,y	Horizontal and vertical position
color	Outline color of points, lines or borders
fill	Interior color for shapes
size	Diameter of points or thickness of lines
shape	Style of a point (0-25)
alpha	Transparency level (0 for transparent, 1 for opaque)
linetype	Pattern of a line (eg. solid, dashed, dotted)
group	Used to identify distinct sets of data

2. Geometric Layers (geom_*)

These parameters are often used outside of aes() to set "static" values for every point or line in that layer

data	Overrides the global dataset
stat	Overrides the default statistical transformation (eg stat= identity)
position	Adjusts how elements sit (eg. position= “dodge” for side-by-side bars)
na.rm	If TRUE, missing values are removed silently
point	Create a scatter plots
line	Create a line graph
bar	Create a bar chart
histogram	Create a histogram
boxplot	Create a box and whisker plot
smooth	To add regression line method=”lm” Fits a linear model (straight line) method= “loess” Fits a Local Regression (curvy line that follows data trends); this is the default for datasets with <1000 points se= FALSE Removes the shaded standard error band around the line span Control how “wiggly” the line is (for loess)

3. Theme Customization (theme)

The theme() function controls non-data elements. It has nearly 100 parameters, categorized by "element functions" 

Text Element (element_text)	plot.title, axis.title.x, legend.text Parameters include family, face, size, hjust and vjust
Line Element (element_line)	panel.grid.major, axis.ticks Parameters include color, linewidth and linetype
Rectangle Elements (element_rect)	panel.background, plot.background Parameters include fill and color
Blank Elements (element_blank)	Removes an element entirely (eg. panel.grid.minor= element_blank()

4. Scales and Labels (scale_* & labs)

Scales translate data into the visual aesthetics you see

name	Axis or legend title
limits	Min/Max values of the scale
breaks	The specific points where labels/tick appear
labels	The text that appears at those breaks
labs()	A shortcut function to set title, subtitle, caption, x and y labels

5. Faceting and Coordinates (facet_* & coord_*)

Faceting	facet_wrap (~variable) facet_grid (rows~cols) Parameters include scales (eg. free_x) and nrow/ncol
Coordinates	coord_flip()       [swaps X and Y) coord_polar()    [circular plot) coord_cartesian ()      [zooming without removing data]