Intent

• Did a test project to create custom/ improved k-means algo.
• Did not use sklearn implementation of algo.
• Create a jupyter notebook to show modeling steps in details.

Initial Start

• This assignment was done using docker container.
• In order to play with Jupyter Notebook, launch docker container with command provided below.
• Start Jupyter container

  docker run --rm --name ds-notebook -v $(pwd):~/work jupyter/scipy-notebook:6d42503c684f

Content Explanation

• custom_kmeans.py: Implementation of kmeans
• extmath.py: Mathematical functions necessary for kmeans
• kmeans_assignment.ipynb: Ipython notebook contains Deliverables and the use of custom kmeans algorithm.
• kmeans_assignment.html: contains Deliverables and the use of custom kmeans algorithm (generated from notebook)
• output..._: files for graphs and csv, generated part of solution