Main.ipynb details the steps taken to merge all LEHD OD data from 2002-2013, convert all of the data to 2010 Census Blocks, and then perform a k-means clustering on the total number of jobs in NYC per OD pair as a time series. Plots of the individual lines for each cluster are shown, overlaid by the corresponding cluster centers.

In addition, randomly generated OD data is clustered and plotted in order to compare it to the results of the LEHD data. It can be seen that the random data lacks the structure of the LEHD data.