Hi there folks! In a world of ever-changing options for transportation, it's evident that the traditional method of driving a car from point A to point B isn't the only thing we have to do anymore. This is evident with the growth of things like ride-sharing with Uber, scooter rentals with Lime, and of course, bicycle rental with Ford GoBike.

In this Udacity project, we'll be taking a look at some data provided to us by Ford's GoBike project. Ford has kindly made this data available to us at this provided URL.

Unfortunately, the dataset in and of itself provides little to be learned from. It contains very few quantitative and categorical features to glean insights from. As I note in the slide deck, there are a number of ways in which this dataset could be augmented, and it appears through searches online that this dataset may have been more robust at one point in time.

Here are a few examples in which I would have hoped to see more information:

• How did repeat users account for the rides in the dataset?
• What impact does the electric vs. non-electric bike types have on this dataset?
• How might knowing some historical weather patterns add additional insights for us?

Still, not all our efforts were futile. We did get a decent understanding at the disparity between rider age and rider membership. Let's highlight a few of those in the next section.

• Gender has no real effect on ride duration. One might expect that males would be taking the longest rides, but the data doesn't indicate this at all. Across all ages, the differences between duration of male riders vs. female & other riders is negligible. In fact, it wasn't uncommon that females and those who marked their gender as "Other" would often ride longer than male riders.
• Folks in their early 30's comprise the bulk of the rides. This one is the one that surprised me the most. Knowing that people as young as 18 can rent one of these bikes, I definitely expected to see the dataset to have a much stronger right skew. While a right skew does exist, isn't extreme given that people in their 30's tend to ride more often than people in their 20's or even in their teens.
• Rides were longer in the summer and shorter in the winter. This one wasn't as surprising, but it is still notable nevertheless. While there was a steady incline in riders over time in general, there were still notable duration spikes in the summer months and dips in the winter months. (Again, this is where it would have been great to have historical weather to validate this to an extent.)

As already shared above, Ford kindly provided us with the data we'll need for this project. Because this data is always fluctuating, I downloaded to my laptop and uploaded to GitHub what I worked from as of April 24, 2019. This way, if you want to get the same results that I did, you can simply leverage the same data that I used!

Here are the fields provided to us within the primary source data sets. We won't necessarily be working with all of these, but it's still good to know what's in here.

 - Trip Duration (in seconds)
 - Start Time and Date
 - End Time and Date
 - Start Station ID
 - Start Station Name
 - Start Station Latitude
 - Start Station Longitude
 - End Station ID
 - End Station Name
 - End Station Latitude
 - End Station Longitude
 - Bike ID
 - User Type (Subscriber or Customer - "Subscriber" = Member or "Customer" = Casual)
 - Member Year of Birth
 - Member Gender

Here's a brief description of all the files at are available as part of this project.

• ford-gobike-trip-data: This folder contains all the data we downloaded and then later assembled / cleaned for use in this project
• ford-gobike-exploration.ipynb: This jupyter notebook is where we do all the exploration of the data to be used in our explanatory analysis and slides later.
• ford-gobike-explanation.ipynb: This jupyter notebook synthesizes the information gleaned from the exploration notebook and will be the foundation for the slide deck down below.
• output_toggle.tpl: Provided by Udacity and the good folks who built nbconvert, this hides the code in my slide deck built off the jupyter notebook above.
• ford-gobike-explanation.slides.html: And finally, this is the slide deck that is built off the jupyter notebook above.

These are the following packages I used throughout this project.

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sb
import os
import requests
from io import BytesIO
from zipfile import ZipFile