Indian Commemorative Postage Stamps: A Data visualization
Information/Data visualization is much awaited topic for one who’s studying Information Design at NID (Bengaluru). After a wait of one full semester and a couple of weeks into second one, we were asked to do a data visualization on a topic of choice.
Coincidentally, for me, I was starting to develop an interest in writing postcard to far-away friends and read about stamps. After a week full of failed attempts at digging for a gold-mine of a topic, It dawned upon me that, Hey, why not visualize Stamps, Indian Stamps at that. As a result, This visualization was created, to explore the types and the growth of Indian Commemorative stamps released from Indian Independence, till 2000.
What are Commemorative Stamps?
A commemorative stamp is a postage stamp, often issued on a significant date such as an anniversary, to honor or commemorate a place, event, person, or object. The subject of the commemorative stamp is usually spelled out in print, unlike definitive stamps which normally depict the subject along with the denomination and country name only.
What is This Visualization about?
The visualization tries to explore the number of stamps released per year, since India’s Independence. It visualizes the growth In different categories, price values.. both annually and Individually and does a comparative analysis among different categories of stamps.
The Visualization
Below is the full blown, A0 Visualization of the stamps (Commemorative). The process for the creation is explained in detail below.
How to Read It?
Zooming In
We start from the foot of the visualization, which demonstrates the year in question and the size of the circle and number below the year is the combined price value of all the stamps released in that year (in Indian Rupees, Paisa and Annas).
We move up along the time line for Category circles. Each category is color coded, and the smaller circles around the category circle are the Number of stamps released in that year.
We then get horizontal lines, arranged according to Months, and each stamp is plotted as a circle(bubble), who’s size is equal to its price value.
At the end of the final month, we have a combined price value for that particular category.
Insights Gathered
Few Interesting Insights that came into light are as below:
Data Sources and Scraping
The data was collected from here and here. As the number of stamps are very large in number, Manual collection of data was out of question. That’s when my three year experience as a python developer came in handy.
I wrote a python script using Selenium and Beautiful Soup, to scrape the above mentioned websites. The collected data was stored in CSV format and later consumed for the visualization.
Nothing comes easy and same can be said for data. As if writing a Python Scraper wasn't hard enough, Cleaning it was equally appalling.
The denomination column in the data, was a string, with price value and currency appended together. Along with that, column had multiple currency values, and everything had to be converted to one common denominator/currency. A quick pandas manipulation did wonders for me.
Story was same with Date column, being a string, and later converted to python date-time format.
D3.js in action
I was aware of tableau but I wasn’t sure how to achieve a form which was in my mind for this visualization. I wanted something… I had full, pixel-level control on. I was learning D3.js, at a moment and thought it was a perfect moment to experiment with it.
I wanted a Tree like structure which was branching/splitting out. Root was an Year and Leaves were categories of stamps. With D3’s stratify and tree methods, I was able to do it.
Rest of the SVG elements were circles and lines and I was able to visualize stamps for a single year successfully. Later, It was a matter of using a quick while loop and generate similar visualization for all the years, from 1947–2000.
The visualization generated for all the years were later refined and adjusted using Adobe Illustrator.
In the sum total of it, I really enjoyed:
- How detailed the data was
- Using Python, Selenium and Beautiful-soup to scrape it
- Pandas to structure it
- Playing with D3.js to generate pixel-perfect visualization
I wasn't ready to let go of the topic just yet, so, I created a stamp of the visualization of stamps, just for fun.
In conclusion, This was my first attempt at Information/Data Visualization and there’s a lot to be discovered in this beautiful field of study. I would love to hear your suggestions, thoughts and ways in which I can improve myself. Thank-you for scrolling all the way.
Below is A0 print mock-up of the same.
