R Language Grammar Of Graphics Concept Complete Guide

Explaining the Grammar of Graphics Concept in R Language

Key Components of the Grammar of Graphics:

1. Data:

   • The input dataset that you want to visualize.

2. Aesthetics (aes):

   • These are visual properties of the objects in our plot, such as position, color, size, shape, and transparency.
   • They map variables in the data to visible features on the plot.

3. Geometry Objects (Geoms):

   • These represent the geometric shapes that will be plotted, like points, lines, rectangles, etc.
   • Examples include geom_point(), geom_line(), geom_histogram(), geom_bar(), etc.

4. Statistical Transformations (Stats):

   • These transform the data before it is plotted, for example, summarizing the data or applying statistical models.
   • Common stats transformations include count, sum, mean, and binning.

5. Coordinate System (Coord):

   • This defines the system where data is mapped into the plot space.
   • Types of coordinate systems can vary (Cartesian, polar, etc.).

6. Facets:

   • These divide the plot into subplots based on categorical variables.
   • Faceting allows us to create multiple plots within one figure to compare different levels of a variable.

7. Scales:

   • Scales control how the raw values of the data are transformed into the aesthetics.
   • Examples include scales for x-axis, y-axis, colors, sizes, and shapes.

8. Themes:

   • Themes control the external appearance of the plot, such as background color, font type and size, gridlines, labels, titles, legends, etc.,
   • They help to focus on the data by reducing unnecessary visual elements.

Detailed Steps to Create Plots Using ggplot2:

1. Library and Data Preparation:

   • Load the ggplot2 library using library(ggplot2).
   • Prepare your dataset. It should generally be in a tidy format with each column representing a variable and each row an observation.

2. Start a Plot:

   • Begin creating your plot with the ggplot() function where you specify your data.
   • Syntax: ggplot(data = data_name), e.g., ggplot(data = mtcars).

3. Add Geometries (Geoms):

   • Use geometry layers to add shapes to the plot.
   • Syntax: + geom_layer_name(mapping = aes(...)). For example, + geom_point(mapping = aes(x = wt, y = mpg)).

4. Specify Aesthetics (aes):

   • Map variables in your dataset to the aesthetic properties within the aes() function.
   • This step connects the data to the plot, determining where the points go (x, y positions), what color they are, their size, shape, and so forth.

5. Apply Statistical Transformations (Stats):

   • Include stat layers if necessary, e.g., for summarizing data or fitting models.
   • Syntax: + stat_function(...), +stat_summary(...), etc.

6. Set Coordinate Systems (Coord):

   • Modify the plot’s coordinate system using functions like coord_cartesian(), coord_flip(), or coord_polar().
   • This can change how the plot looks spatially, for instance flipping axes or moving to polar coordinates for circular graphs.

7. Create Facets:

   • To create small multiples of the same plot for different categories in your data, use facets.
   • Syntax: + facet_wrap(~ variable), + facet_grid(variable1 ~ variable2).
   • This is useful when you want to compare distributions or trends across several groups.

8. Modify Scales:

   • Adjust the scales used for each aesthetic with functions like scale_x_continuous(), scale_color_gradient(), etc.
   • This step customizes how the variables are visualized, providing labels, limits, breaks, and other formatting options.

9. Enhance with Themes:

   • Finally, customize the overall theme of the plot using theme() function.
   • This includes things like the plot's title, axis labels, plot margins, legend properties, etc.
   • Syntax: + theme(...).

Example Plot in ggplot2:

Let’s assume you’re interested in plotting mpg vs. wt for the mtcars dataset to see the trend in miles per gallon (mpg) against car weight (wt).

library(ggplot2)

# Start plotting with data and initial geometry layer
ggplot(mtcars, aes(x = wt, y = mpg)) + 
    # Add point geometry layer 
    geom_point(color = 'blue', size = 2.5) +
    # Fit a linear model to the dataset
    geom_smooth(method = 'lm', se = FALSE) +
    # Set the x-axis range, break points and labels
    scale_x_continuous(limit = c(2, 5), breaks = seq(2, 5, by = .5)) +
    # Set the y-axis continuous scale with labels
    scale_y_continuous(labels = scales::comma) +
    # Customizing the plot with a theme
    theme(plot.title = element_text(hjust = 0.5),
          axis.title.x = element_text(color = 'red'), 
          axis.title.y = element_text(color = 'red'),
          panel.background = element_rect(fill = 'lightyellow')) +
    # Adding titles, subtitles, and captions
    labs(title = "Relationship Between MPG and Car Weight",
         subtitle = "Data from the mtcars dataset",
         caption = "Source: R built-in datasets",
         x = "Weight (1000 lbs)",
         y = "MPG")

In this example:

• The dataset mtcars is passed to ggplot() function.
• aes(x = wt, y = mpg) maps weight to the x-axis and mpg to the y-axis.
• geom_point() adds scatter points indicating individual car data.
• geom_smooth(method = 'lm') overlays a line representing the fitted linear model.
• Axes are customized with scale_x_continuous() and scale_y_continuous().
• theme() adjusts the entire plot's look.
• labs() provides title, subtitles, and labels.