Data Visualizations and Python: A Match Made in Heaven

My dear reader, how are you? السلام عليكم

Visualization and belief in a pattern of reality activate the creative power of realization – A. L. Linall Jr.

Creativity can be defined as a capacity of a being to discover and unleash hidden realities in complex structures and shattered pieces of information. Visualization (Viz) acts as a catalyst for creativity. And, Python provides great Viz-tools to make the process of establishing the hidden patterns of data to reality.

In this post, I provide a road map and list a few key tools for you to learn the art of data Viz using Python programming language and master your skills as a data analyst.

Preliminaries: Why Python? I answered it in one of my previous posts DirectMe. New to Python? learn it in a structured way DirectMe. Not good at data analysis? A previous post of mine on Python for data science can help you get started DirectMe.

Enough with motivation? Let us get started then.

Where to start learning data visualizations? I recommend starting in a structured way. Don’t like to follow structures and rules? Me too! But, you have to unless you become proficient and know enough about breaking the existing structures and make them better. So, start with a free introductory course on coursera by IBM DirectMe. Revise your concepts by taking another free course at pluralsite designed by YK from CS Dojo DirectMe.

After these two courses, you may get a basic understanding of how to visualize data using matplotlib, i.e., one of the most popular, open-source plotting library, and a widely used stable tool. There are several tools built on top of matplotlib that makes the Viz process easier by reducing the efforts of data analyst for customizing the plots. One of such extension includes seaborn tool. Need a quick Introduction? DirectMe. Other extensions include ggpy, yellow brich, scikit plot (I love this one too), pandas, basemap, networkx.

So that’s it? well, no. These plots lack something. That is the facility of making the visualizations interactive to be used in web applications. The answer is Javascript. There is a tremendous amount of support on interactive data visualizations provided by javascript based technologies over the years. Few popular and widely used javascript based viz-tools include plotly, toyplot, bqplot, ipyleaflet, ipyvolume and cufflinks.

In life, as you grow old, you get a lot of experiences. Similarly, everyone around you has their own experiences. Consider these experiences as pieces of a puzzle. When you interact with your fellow and get involved in an exercise of sharing experiences, those pieces of the puzzle combine and that is where the magic happens. Yes, you got it right. How about combining the powers of data visualizations from Python and JavaScript-based tools?

There is a class of potential software tools which comprises of a blend of the two underlying technologies (Python and JavaScript) and take the art of data visualization to a far next level. These include mpld3, d3js, d3po, Vega, Vega-lite, Altair, and Vincent.

Now, you have a great insight into the landscape of data visualization tools which are available on the web for you to explore. So, Happy Hunting 😉

All of the tools, I mentioned in this post are open-source and free for you to play with. Stay tuned for the future posts and please do not hesitate to suggest any other useful Viz-tool that I might have missed.