Part $1$: Data Collection and Preparation

You're given a Python script $($us$\_$election$\_$results$\_$kmeans.py$)$ that extracts a

table from Wikipedia, listing the U$.$S$.$ presidential election results by state from

$1972$ to $2020.$ This table is then cleaned and formatted into a DataFrame.

Familiarize yourself with pandas in Python in order to manipulate the data

frame for the following questions. The script also gives you step by step details

and hints on how to do every question in this assignment. You will need to

understand how Pandas and KMeans in sklearn works.

Task:

Use the provided code snippet to collect and prepare the dataset.

Convert all $"$D$"$ characters in the dataset to $0\text{'}$s$,$ and all $"R"$ characters

to $1\text{'}$s$,$ making the data numerical for analysis. Hint: Google the replace

method in pandas.

Print the first five rows of the cleaned dataset to ensure it has been cor$-$

rectly processed. Hint: Use the head method in pandas.

Part $2$ : Data Subset Identification

Tasks:

Identify the states that have only voted Republican in the given time

period.

Identify the states that have only voted Democratic in the same period.

Find the list of states that voted exactly the same as Illinois over this time

period.

Discussion: Briefly comment on your findings, noting any interesting pat$-$

terns or anomalies.

Part $3$: $K-$means Clustering Analysis

Tasks:

Finding the Optimal Number of Clusters $(K)$ : Perform $K-$means cluster$-$

ing on the dataset for multiple values of $K($at least $5).$ For each $K,$ try

a few $($at least $5)$ different random starting points to ensure stability.

Plotting Within$-$Cluster Variance: Create a plot with the number of clus$-$

ters $K$ on the $x-$axis and the total within$-$cluster variance on the $y-$axis.

This will help you visually determine the optimal number of clusters. Read

about the elbow method on the internet.

Choosing $K$ : Write a sentence justifying your choice of $K$ based on the

plot.

Part $4$: Interpretation of Clusters

Once you have chosen an optimal value for $K,$ provide an interpretation for each

cluster identified by your $K-$means analysis. Consider the political leanings and

consistency in voting patterns across the states within each cluster.

Note: For the implementation of $K-$means clustering, you may use the

KMeans class from the sklearn.cluster module. This exercise assumes famil$-$

iarity with Python, pandas for data manipulation, and scikit$-$learn for machine

learning tasks.