Sinaplot vs Violin plot – Why Sinaplot is better than Violinplot in R

In this article, we are going to learn sinaplot and violin plots, and compare them in R programming language.

Sinaplot and violin plots are both useful visualization tools in R for displaying distributions of data. However, sina plots have some advantages over violin plots that make them a better choice in certain situations. In this overview, we will compare sina plots and violin plots and explain why sina plots may be a better choice for visualizing data in R.

What are plot functions and why they are needed?

Plot functions in R are used to create visual representations of data, which can make it easier to understand and analyze the data. Some of the main reasons why we use plot functions in R include:

1. Exploration: Plotting data is an efficient way to explore the characteristics of a dataset, such as distribution, patterns, and outliers. It can also help identify any issues with the data, such as missing values or errors.
2. Communication: Plotting data is an effective way to communicate the results of an analysis to others. Visualizations can be more easily understood than raw data and can be used to convey complex ideas in a simple and intuitive way.
3. Model Evaluation: Plotting data can also be useful for evaluating the performance of a model, such as a statistical or machine learning model. It can help to identify any patterns or trends in the data that may not be captured by the model, which can be used to improve the model’s performance.
4. Decision Making: Plotting data can help to make better decisions by providing a visual representation of the data that can be used to identify trends and patterns. It can also be used to identify areas where further analysis is needed.
5. Identify outliers: Plotting data can help to identify outliers in the data which could be caused by measurement errors or data entry errors.

In summary, plot functions in R are a powerful tool for visualizing data and can be used to explore, understand, and communicate the results of an analysis in a clear and intuitive way. R has a wide variety of plot functions available, which can be used to create different types of plots depending on the data and the analysis performed. In this article we will only learn about two plots i.e., sinaplot() and violinplot().

Violin plot

Violin plot is a combination of a box plot and a kernel density plot. It shows the distribution of the data across all levels of a categorical variable by plotting a violin-shaped figure for each level. The violin plot shows the density of the data using the width of the violin, and the box plot shows the summary statistics of the data using the box and whiskers. Additionally, the violin plot shows the data distribution by kernel density, a smooth estimate of the probability density function of the data.

The violinplot() function is part of the “ggplot2” package, which can be installed using the following command:

install.packages("ggplot2")

Syntax:

ggplot(data, aes(x = x_variable, y = y_variable, fill = group_variable))+geom_violin(trim = TRUE/FALSE, draw_quantiles = c(0.25, 0.5, 0.75))

Parameters:

• data: name of the data frame.
• aes(x = x_variable, y = y_variable): variables for the x-axis and y-axis.
• fill = group_variable: variable to group the data by and fill the violin with different colors.
• trim: whether to remove the tails of the violin plot.
• draw_quantiles: quantiles to display on the violin plot.

Examples 1:

In this example, we have created three groups of data using the rnorm() function. Then we created a data frame with the data and group information. Then we created a violin plot using the ggplot2 package, passing in the data frame and specifying the x and y variables and the group variable. We can also add a title, x and y labels, and also specified the geom_violin() function to create the violin plot. We’ve used trim = FALSE to show the complete violin, and draw_quantiles = c(0.25, 0.5, 0.75) to show the quartiles in the violin.

R

# Import required library
library(ggplot2)

# Create some example data
set.seed(123)
data1 <- rnorm(50, mean = 5,
               sd = 2)
data2 <- rnorm(50, mean = 8,
               sd = 1)
data3 <- rnorm(50, mean = 10,
               sd = 3)

# Create a data frame with the data
# and group information
data_frame <- data.frame(x = c(data1,
                               data2,
                               data3),
              group = rep(c("Group 1",
                            "Group 2",
                            "Group 3"),
                           each = 50))

# Create the violin plot
ggplot(data = data_frame, aes(x = group, y = x,
                              fill = group)) +
              geom_violin(trim = FALSE,
              draw_quantiles = c(0.25, 0.5, 0.75)) +
              ggtitle("Violin Plot Example") +
              xlab("Group") +ylab("Value")

Output:

Example 2:

In this example, we create a violin plot that compares the distribution of miles per gallon for vehicles with different numbers of cylinders. The plot shows the density of the data using the width of the violin, and the box plot shows the summary statistics of the data using the box and whiskers. Additionally, the violin plot shows the data distribution by kernel density, a smooth estimate of the probability density function of the data..

R

# Importing library and data set
library(ggplot2)
data(mtcars)

# Creating a violin plot
ggplot(mtcars, aes(x = factor(cyl),
  y = mpg, fill = factor(cyl))) +
  geom_violin(trim = TRUE,
  draw_quantiles = c(0.25, 0.5, 0.75)) +
  geom_boxplot(width = 0.1,
               fill = "white") +
  ggtitle("Violin Plot of Miles per Gallon by Number of Cylinders") +
  xlab("Number of Cylinders") +
  ylab("Miles per Gallon")

Output:

Sinaplot

Sinaplot is a variation of a violin plot. It shows the distribution of the data across all levels of a categorical variable by plotting a sine wave shape for each level. The sina plot shows the density of the data using the amplitude of the sine wave, and also shows the data distribution by kernel density. The sina plot uses a sine wave shape to represent the distribution of data, rather than a traditional violin shape. This allows for a more compact representation of the data, which can be useful when working with large datasets or datasets with many overlapping points.

To plot sinaplot in R, we need to have “plyr” package and “sinaplot” package. Execute the below commands to install them.

install.packages("plyr")

install.packages("sinaplot")

Syntax:

sinaplot(x, groups, …)

Parameters:

• x: a numeric vector of values
• groups: a factor or character vector defining the groups to be separated in the sinaplot
• …: additional arguments, such as col, pch, xaxt, ann, bty, etc.

Note: Download the Iris data set.

Example 1:

In this article, we are going to create a sinaplot graph by using the sepal  length on y-axis and species on x-axis using in iris data set.

R

# Import required library
library(sinaplot)

# Load data set
data("iris")

# Plot graph
sinaplot(Sepal.Length ~ Species,
         data = iris,
         pch = 20,
         col = rainbow(3),
         ann = FALSE,
         bty = "n")

Output:

Example 2:

In this example, we creates a sinaplot using the “airquality” built-in dataset in R. The sinaplot() function from the “sinaplot” library is used to plot the relationship between “Ozone” and “Wind”.

R

# Import required library
library(sinaplot)

# Load dataset
data("airquality")

# Plot graph
sinaplot(Ozone ~ Month,
         data = airquality,
         pch = 20,
         col = rainbow(5),
         ann = FALSE,
         bty = "n")

Output:

Key Differences between violin plot and sinaplot.

1. Dimensionality: A sinaplot is a 3D scatter plot while a violin plot is a 2D plot.
2. Purpose: A sinaplot is used to visualize the distribution of a continuous variable when the data has a high degree of scatter, while a violin plot is used to display the distribution of a continuous variable on a single dimension.
3. Visual representation: A sinaplot shows the distribution of the data in all three dimensions, showing the relationship between the x, y, and z variables, while a violin plot shows the distribution of the data in a single dimension.
4. Display of the data distribution: A sinaplot shows the distribution of the data in three dimensions, while a violin plot shows the entire distribution of the data.
5. Smoothness: A sinaplot is a scatter plot, showing individual data points, while a violin plot is a smoothed version of a histogram.
6. Shape: A sinaplot is represented in a 3D space and can have any shape, while a violin plot is represented in a 2D space and has a violin-like shape.
7. Outliers: A sinaplot can show outliers more clearly as it is a scatter plot showing individual data points, while a violin plot doesn’t show the outliers clearly.
8. Plotting method: A sinaplot is typically plotted using a 3D scatter plot and rotation on the x-y plane, while a violin plot is plotted using a smoothed version of a histogram.
9. Complexity: A sinaplot is more complex to interpret as it shows three dimensions of data, while a violin plot is relatively simple to interpret as it shows one dimension of data.
10. Additional information: A violin plot can provide additional information such as median, mean, and kernel density estimates, while a sinaplot doesn’t provide this information.

Why Sinaplot is better than Violinplot in R?

• Both violin plots and sina plots are types of probability density plots that can be used to visualize the distribution of a dataset. However, there are a few key differences between the two that may make one more suitable for a particular use case than the other.
• One advantage of sina plots over violin plots is that they are less prone to overplotting. This is because sina plots use a sine wave shape to represent the distribution of data, rather than a traditional violin shape. This allows for a more compact representation of the data, which can be useful when working with large datasets or datasets with many overlapping points.
• Another advantage of sina plots is that they can be more effective at showing the underlying distribution of a dataset. This is because the sine wave shape used in sina plots is able to capture more subtle features of the distribution, such as skewness or kurtosis, that may not be as easily visible in a violin plot.
• Finally, sina plots also provide a more intuitive way to compare the distributions of multiple datasets. This is because the sine wave shape used in sina plots allows for a more direct comparison of the amplitude of the distribution between different datasets.
• Violin plot are more common and are more easily interpretable for people who are not data scientists. They also have more options for customization, such as adding additional information like box plots and showing the number of observations in each category.
• Ultimately, the choice between using a violin plot or a sina plot will depend on the specific requirements of your analysis and the characteristics of your dataset.