On a pedestrian workday, one of my colleagues mentioned how excited they were to use “Tableau” to present some analysis to their managers. Intrigued by this mysterious French tool, I asked if I could see what they were working on, and was instantly attracted to the ease with which I could drag and drop “pills” into different areas and create charts that would took me hours to make in Excel. It triggered my traumatic experiences from making analyses for internships in college with sample data and basic color-coded line graphs.
Understanding the basics of columns (or “fields”) and rows (or “records”) that make up databases was a headache at first, but once I grasped those two concepts alone, my learning accelerated dramatically. Looking back at this time, I really didn’t understand what data was, and how it fit into a database. With this fundamental understanding, however, I could grasp basic SQL queries and visualize what the results might look like.
In the year that followed, I read everything I could get my hands on about Tableau in particular. My co-worker friend shared a book called Practical Tableau by Ryan Sleeper, which is an amazing resource to get started, and to refer back to. I used real data from work, Tableau Public, Kaggle, and data.world to follow Ryan’s instructions for creating different types of charts. My first moment of pride was making a “wheel” chart like Ryan suggests, and presenting my first dashboard at my job.
After grasping the basics, I jumped at any work opportunity to do some analysis and dig up answers from our database, which further boosted me along the learning curve, and resulted in more data-related projects. The project requests started getting more complicated, so I learned more out of necessity. Now I had to know which types of charts lend themselves to certain data types, or data questions. I needed better SQL skills to pull and transform messy data into something that can be easily used in Tableau’s interface, and I needed to figure out how to code enough to write calculations to make my analyses more insightful.
To delve deeper into data science, I signed up for Codecademy and have been using it consistently for two years. It’s been an effective tool for practicing data science skills like SQL and the Python libraries that allow for data cleaning and wrangling, plotting graphs, and advanced skills in machine learning.
Currently, I’m competent in using python libraries to create and manipulate data frames, build visualizations, deploy machine learning algorithms, and make basic neural networks for deep learning applications such as natural language processing. These concepts were completely foreign to me just two years ago, and I am excited about how much there still is to learn.
The best part about this whole process is the community of learners. Everyone seems to feel like a beginner, because there is no way to master every single thing in the data science world. There are so many things to learn that it’s best to work with what we know so far, and pick up knowledge as we explore projects. We’ve all heard that the best way to learn is by doing, and data science is not an exception.