Blog 1

Data has become a buzz word and rightly so. Data science and analysis is making waves in every field be it banking, technology, sports, medicine. However, simply analyzing the data isn’t enough. The insights need to be presented in an understandable format.

For eg. After analyzing using Python I get tabulated data as numbers. These numbers don’t give us a solution yet and are difficult to relate to the problem.

How can these numbers be used to give proper insight? Say hello to Visualization!

Visualizations act as a campfire around which we gather to tell stories.
―Al Shalloway

As a Computer Science Grad student in my third semester, I had already taken the core required courses; Algorithms, Data Structures, etc. This semester I wanted to do something creative and learn more visualization skills. Data Visualization under Prof. Arlene Ducao was a great suggestion and I am glad I went ahead with it.

The first lecture entailed history of visualizing data. I liked how the interactive session focused on how shape, size, color can help make better representations. Things that I found interesting:

  • The ancient Chinese seismometer of AD 132 Han dynasty 
  • Visualizing the effect a bullet has on an apple in a single second
  • Different ways in which calendars and timelines were maintained throughout history

The later half included a fun session where we formed groups to visualize data we ourselves gathered to questions we formulated.

Our question: Food preference based on country of origin

Data collected: Name, most favorite cuisine, country of origin

Materials used: Self-made map, u clip

Below are images of our key for each type of cuisine and the colored u clips placed on correct country.

 

Key

Map

Through our visualization we inferred 

  • Citizens of North America prefer Japanese cuisine
  • Most citizens of India prefer home-made Indian cuisine

The challenges we faced were typically in data collection and decision making. We could have inferred more accurately with a bulk of data as compared to just the class data, but it was a good attempt with whatever as at hand. I really enjoyed the different flow of this class!