Power Chart of Chinese Provinces

Economist just posts an interactive visualization Chinese Equivalents on their website. It’s a very interesting approach. (Somehow I feel it has an psychological side-effect by saying one province is equivalent to France while it’s neighbor is equivalent to Kenya, though noted in terms of population.)

I got curious how we can visualize how actually important the Chinese provinces are. I was reading Gastner & Newman’s paper Diffusion-based method for producing density-equalizing maps (PNAS 2004) at the time. So I set out to make my first density-equalizing map.

Newman’s code on his website deals with raster image only. I also tried to implement a diffusion simulator in Processing, but it was hard to preserve all the details of a vector map. Below is my first shot. It took (quite) some manual effort to remove the bad points. Still seeking solutions. Anyway, enjoy.

You can recognize in this map how unbalanced China is – the west is barely occupied due to challenging natural environment, and population keeps flowing from the middle towards the economic centers (Beijing and the southeast coast). They become both productivity and pressure for big cities.

What about looking at the provinces from a social network’s perspective? In the following graph I measured how often each two provinces appear in the same media coverage. It is clear now that Beijing is the absolute, mono-center of all China. The social power is not proportional to a province’s population. The south and the east coast get far more attention than inland provinces, which is a sad fact.

Tools used: Processing, Tulip, Illustrator


2 Replies to “Power Chart of Chinese Provinces”

  1. very impressive and inspiring. i have been following your mapping work. i also tried to guess the codes/procedures behind your graphics like “connected states of america” and start learning softwares listed below those graphics. many thanks.

    could you also kindly let me know how did you find the the social power matrix? thank you very much indeed.

  2. Hello, I would like to know how your data set was prepared and then how was it overlaid to your map? I have a number of Fire Departments that interact with each other and would like to show how they are involved by incident count throughout the course of a year. I’m assuming you used lat/lon data to place the nodes appropriately for the cities?

    Thanks for any insight or advice you could provide…Don

Leave a Reply

Your email address will not be published. Required fields are marked *