The biggest challenge was to generate an intresting tree map without getting something similar to the most popular D3 TreeMap example. I used google charts but the data needed alot of reformatting. I used open refine to change the json format to a 2 column format understood by google Docs. The second problem I faced was that google charts doesn't consider case senstivity when handling data, so I had to rename alot of fields to solve the mislinkage.


The previous image shows the initial plot of my Facebook network. It seems very interconnected to the level of not getting any output from it.

In the previous plot I was able to group them and spot members that I worked with in a company called POET, still not much to get.

I was also able to group them into 5 groups: Family, Friends, Co-workers in ISFP, Co-workers in POET, and the rest. It was intresting to see that all my groups are inter-related. To my surprise some of my co-workers were friends with to my wife and other family members.

In the previous plot I tried to explore the relation between my family members and my other friends from different groups.

What I found didn't like about NodeXL, it is not easy to explore and the documentation is not very helpful. I wanted to make group with multiple parameters for example I wanted to see who in my network is (single, married), and (male, female). I wasn't able to have this classification easily.


This template downloaded form free website templates