Good Evening Followers,

This week in Data Visualization Class we focused on building visualizations and dashboards which effectively answered the requirements of a project outlined by our teacher. In class, our professor presented us with a dataset of Uber pickup data and 2-3 challenges described in the project overview.

The first challenge was to build individual dashboards comprised of several visualizations which individually answered questions like: How many total uber pickups happened in each borough of NYC? What were the total monthly pickups for each borough? What were the total pickups per holiday? What were the average daily pickups numbers for each borough?

This is the dashboard I built, comprised of four different visuals which answered those questions.

Our next challenge was to refine this dashboard in our project teams using all of our different ideas and individual dashboards. Our team decided that bar charts were objectively the best for representing this kind of data, especially in the ways that the requirements asked for the data. Specifically, we decided that using the stacked bar graph format, using categorical-style coloring methods (a method we learned this week in data viz class) was the best way to break up the different boroughs of NYC that we were analyzing.

Additionally, we used the quantitative labels on top of each bar graph for the total pickups for each borough to know the precise numbers. We decided for the rest of the data that this was unnecessary because knowing the ballpark numbers were all you need to know since this data was unique and could change if given more data over more time.

Lastly, our professor challenged us to create one visualization which attempted to answer all of the requirements of the dashboards we previously built. Additionally, he asked us to attempt to incorporate the weather data to see how we could incorporate it into our visuals.

We decided to use this geographical visualization because we believed it best used sequential coloring methods to demonstrate the density of pickups across the different boroughs, using different time-related filters to show how the intensity of the colors changed. It was also the easiest to use because we believed it best incorporated a hover visual to drill down on all of the pertinent data to answer the questions, including the weather data, which we were able to include a visual inside this visual in this hover viz. Once again, we used categorical colors for different weather data.

In review, we were able to take advantage of our newly learned and applied skills of using color in situations appropriately. Also, we strengthed our Tableau skills. Lastly, we were evaluated by a guest speaker, who made unbiased recommendations and reviews of how we presented our data and insights. Insights like how weather temporarily reduced traffic, until it reached a plateau where it didn’t decrease anymore. Or, understanding how we could gather more data to better understand the smaller areas, because we acknowledged how Manhattan data skewed our interpretations of the data.

More to come next week about how we interpret and bring analysis from these visualizaitons. For know, we learned how to use color and best compress our data into one meaning vizual.

 

Signing off for now,

Chandler