cluster

cluster, given a set of data on eruptions of Old Faithful, look for natural groupings. The K-means algorithm can be used to identify locations near which many data points are gathered.

Lecture notes:

• cluster.pdf

• faithful_data.txt, the original 272 pairs of eruption and wait times for a geyser.
• faithful_normalized.txt, the normalized data.
• faithful_standardized.txt, the standardized data.

• data_demo.py, generates, analyzes, and plots several types of data.
• faithful_wait.py, make a histogram of the waiting times.
• faithful_both.py, plot eruption and wait times together, rescale the data, and write to new files.
faithful_guess.py, guess two centers, then cluster the data.
• faithful_lloyd.py, perform four iterations on Lloyd's clustering algorithm.
• faithful_kmeans2.py, cluster the data with the scipy kmeans2() function.
• faithful_energy.py, plot the change in cluster energy as we increase the number of clusters.
• blobs_energy.py, plot the change in cluster energy as we increase the number of clusters for a set of "blobs" data.