What makes a good coder into a great data scientist?
An opinion piece of an up and coming Data Scientist. All feedback is welcomed in the comments. Let me know what you think are the best attributes a data scientist can have!
Data Visualization
Some people can understand numbers very well, and others do better with clearly worded explanations, but just about everyone can ‘get the picture’ from a good visualization. Bar charts, color-coded maps, and even line graphs have become a staple of explaining data through internet media and televised news. Great example libraries to check out for useful data visualizations would be MatPlotLib, Seaborn, Plotly, and Altair.
Data Storytelling
Explaining the inner workings of a machine learning model in a way most people can understand is a talent that is beyond priceless. It can be difficult enough to do all the data wrangling, data cleaning, predictive modeling, hyper-parameter tweaking, and data visualizations, but to then have the ability to completely humanize your explanation of all of that for presentation to people who may have no understanding of how these things work, and also the work that it takes to achieve them, are skills that can separate a great data scientist from the pool of people who are good at writing code. The best library to help unlock the mysterious ‘black box’ of machine learning models in a presentable, explainable manner that I have come across is Shapley.
Interpersonal Communication
In my life since leaving college at the University of Delaware in 1994 to pursue a career as a full-time raver and professional slacker, I have propelled myself into many different social circles and have yet to find one that I wasn’t able to fit in and also embraced. This includes, but is not limited to, clubs, jails, retail work establishments, arts and music festivals, light industrial warehouses, special needs education, food service, forex trading groups, health insurance companies, gaming communities, home improvement roles… My point is, each of these different walks of life has helped me to understand how to communicate with many different types of people in many different ways. I believe that is one skill that sets me apart from the rest of my peers.
Now, I don’t suggest anyone take a 20-year break from their education just to gain life experience, but recognizing that you may have to be capable of accepting and communicating with many types of people from varying backgrounds can only help your value as a data scientist.
Teamwork Skills
Although some companies may only employ a single data scientist. I believe it is more often than not that the opposite is true. From what I have heard and seen in my initial job searches, most companies that have a data scientist want at least 2. Senior and Junior are often titles attached to data science roles. Relying on a single employee to handle all data science tasks would not only be a bad model for sustainability, but it may also put an excessive amount of pressure on a single employee to solve every problem and predict future trends. For these reasons alone it is a good idea to have experience working along with other data scientists. Even for companies that do only have a single role for Data Science, though, you are more than likely going to be working in tandem with a team of front-end and/or backend developers. Learning to work on projects with other humans that may not even speak the same (coding) language with which you are familiar, is essential to aid in overcoming technical issues and solving any number of brand new problems that may arise.
A Healthy Appetite for Knowledge
Although there are surely more attributes that can help mold a great data scientist out of a good coder, I can’t think of a better skill on which to end. No matter how much I learn in the rest of my journey towards data science as a career, I have little doubt that education will be a constant in this field of work. As new technologies emerge that increase the processing speed of computing, also will new tools that test the limits of those technologies. Natural language processing has grown from simply predicting which word you were trying to spell on your flip phone to actually being able to listen, transcribe, and predict groupings of words input verbally to our devices. That shift has happened in a very short span of time. Neural networks are emerging and the always somehow eerie sounding “AI” is already more of a reality than some people realize. A hunger to stay current with a working knowledge of the best tools at their disposal will surely be a valuable trait to exhibit for an up and coming data scientist.
Outro
Thank you for the read! Please share your ideas in the comments about what is the most valuable trait you think a data scientist should have. As always all claps are dedicated to the universe for keeping me intrigued. The universe is, after all, comprised of data, just waiting to be observed, described, and appreciated. Cheers!
