syyunn / kmeans

Implements K-means for 2-dimensional Data

KMeans-clustering

About

This repo implements simple k-means clustering.

Test

conda env create -f environment.yml
conda activate kmeans
python test.py -n_clusters 5 -n_points 100 
# After the above commands, look into the `result` dir to check the generated result

TimeFrame

For algorithmic implementation : 45min
For visualization codes : 1 hour

Code Explanation

• test.py generates -n_points number of data points for -n_cluters number of 2 dimensional multivariate gaussian distributions. The parameters for each distribution, mean and covariance, is random-sampled but cross-dimensional variances are set to be 0 for its visual tidiness.

• The generated data points and its sampled gaussian distributions will be visualized under the name of result/raw.png and result/gaussian.png respectively.

• After data generation, the KMeans runs on the data and its labels and centorids will be plotted under the name of result/kmeans.png

• To compare the Gaussian and KMeans together, please refer to the result/final.png which concatenates every generated pngs.

• To check the convergence, please refer to result/log.txt which logs the intertia and center shift.

TODO

• Taking pictures of shifting centers while KMeans iterates over the process.