##Current Bugs

• FeatureGen outputs a BS row in the first line (see repeated hour value)

##Latest Fixes

• Hours that contain no tweets have appropriate value for time of day (24hr)
• Added functionality to create train/test set via list of indices

##Todo

• Add new features for Part 3 in Feature.py
• Clean up TweetStats.py modules (separate FeatureGen and Part 1 stats)
• Prune DataFrame based on time window (for Part 4) e.g date < Feb 1st
• Fix the feature building (Tweet count should be for next hour)