Module 9: Comparing Base R, Lattice, and ggplot2

 

    For this assignment, I used the built-in iris dataset in R. I chose this dataset because it includes several numerical variables and a grouping variable, Species, which made it useful for comparing different visualization systems. I created plots using base R, lattice, and ggplot2 to see how the syntax, workflow, and final output differed across the three systems.

Base R Graphics

    For the base R portion, I created a scatter plot showing the relationship between sepal length and petal length. Base R was straightforward to use for a basic graph, but it required more manual work, such as adding the legend separately.  


Lattice Graphics

    For the lattice portion, I used a conditioned scatter plot to separate the data by species. This made it easier to compare patterns across groups because each species appeared in its own panel.



 

ggplot2 Graphics

    For ggplot2, I created a scatter plot with trend lines for each species. Out of the three systems, ggplot2 gave me the most polished-looking graph with relatively little code once the structure was set up.


Discussion

    The syntax and workflow were different across all three systems. Base R felt the most direct because I could quickly create a plot with one function, but I had to manually control extra details like legends. Lattice used a formula style and made it easier to separate data into panels, which was useful for comparing species. ggplot2 was the most layered system, since I had to think about aesthetics, geoms, and labels separately.

    Out of the three, ggplot2 gave me the most publication-quality result with minimal code once I understood the structure. It was easier to add trend lines, colors, and labels in a clean way. Lattice was also very useful for grouped comparisons, while base R was the fastest for simple plots.

    The biggest challenge was switching between the different syntax styles. Base R, lattice, and ggplot2 all approach graphing differently, so I had to adjust the way I thought about building each visualization. Once I practiced each one, it became easier to see the strengths of each system.

 

Comments

Post a Comment

Popular posts from this blog

Module 5: Determinant and Inverse of Matrices in R

Image
 This assignment practiced creating matrices in R and computing a determinant and inverse (when possible).   Step 1: Creating Matrices A and B   I created the matrices using matrix() :    Matrix A contains the numbers 1 to 100 arranged into 10 rows. Matrix B contains the numbers 1 to 1000 arranged into 10 rows.    Step 2: Checking the Dimensions To check the size of each matrix, I used: Step 3: Finding the Determinant of A I used the det() function to find the determinant of A: This result shows that matrix A is singular , which means it does not have an inverse.    Step 4: Finding the Inverse of A I attempted to find the inverse of A using:   Result: R returned an error saying that the system is computationally singular. This confirms that A does not have an inverse because its determinant is zero.   Step 5: Testing Matrix B Since B is not a square matrix, I tested:     Result: Both commands returned errors. This...
Read more

Creating and Analyzing Data Frames in R

Image
Intro In Module 3, I worked with data frames in R to organize and analyze polling data from a fictional 2016 presidential election dataset. The purpose of this exercise was to practice creating vectors, combining them into a data frame, and performing basic comparisons across variables using R. The dataset consists of three variables: Name – the candidate’s name ABC – polling results from the ABC source CBS – polling results from the CBS source Each row in the data frame represents one candidate, and each column represents a variable. This structure makes it easy to compare values across different polling sources.   Results and Observations After creating the data frame, I compared the results from ABC and CBS. Some candidates showed higher support in CBS polls, while others performed better in ABC polls. This highlights how different polling sources can produce varying results even when measuring the same data. I also added calculated columns to examine the difference...
Read more

Module 11: Debugging and Defensive Programming in R

  For this assignment, I debugged a function that was supposed to identify rows in a numeric matrix that were outliers in every column using the Tukey rule. I first reproduced the bug by creating a test matrix with random values and running the original function. When I ran the function, I got a warning showing that the logical expression was not being handled correctly. The issue was caused by using && inside the loop. In R, && only checks the first element of a logical vector, but this function needed an element-wise comparison across all rows. Because of that, the code was not correctly updating the outlier matrix. To fix the problem, I replaced && with & . The & operator compares vectors element by element, which matches the purpose of the function. After making that change, the corrected version ran without warnings and returned a logical vector of length 10. I also added defensive programming checks at the top of the function to make sur...
Read more