Dinesh Kumar Paladhi PaladhiDinesh

###Design Choise made : SuperImposed Line charts

I have chosen Olympics dataset. In this dataset I wanted to know how count of men and women participants changed over the years. I also wanted to know change occured in the participants count over the years. In order to do that I felt Multiple line charts are a great way to do. So In the graph I have taken Years on the x-axis and count on the Y-axis. Super Imposition is not seen clearly because the values did not overlap. Men, Women and Participant Line graphs are represented with different colors and the name for each line is given at the end of the line. I have removed x-axis from the line chart because I feel it as an obstacle. I learned about removal of the x-axis from the article "Line Charts : Where to Start ?"

Using D3, create a graph of the Karate club before and after the split.

• Weight the edges with the data from: http://vlado.fmf.uni-lj.si/pub/networks/data/ucinet/zachary.dat

• Have the transition from before/after the split occur on a mouse click. This is a toggle, so the graph will go back and forth beween connected and disconnected.

Take the Twitter graph you generated in question #1 and test for male-female homophily. For the purposes of this question you can consider the graph as undirected (i.e., no distinction between "follows" and "following"). Use the twitter name (not "screen name"; for example "Michael L. Nelson" and not "@phonedude_mln") and programatically determine if the user is male or female. Some sites that might be useful:

https://genderize.io/ https://pypi.python.org/pypi/gender-detector/0.0.4

Use D3 to visualize your Twitter followers. Use my twitter account ("@phonedude_mln") if you do not have >= 50 followers. For example, @hvdsomp follows me, as does @mart1nkle1n. They also follow each other, so they would both have links to me and links to each other.

To see if two users follow each other, see: https://dev.twitter.com/rest/reference/get/friendships/show

Attractiveness of the graph counts! Nodes should be labeled (avatar images are even better), and edge types (follows, following) should

###Insight gained:

• We can order either by region or alphabetical order or number of links.
• By ordering the states in alphabetical order, we can easily track down a state and its number of connections if we traverse through the particular row or column
• By ordering the number of links we get the state with higher number of links on top left and least number of links on bottom right.
• In all the orderings the states from the same region are with the same color.

###Insight gained:

• The state that is connected to many states can be seen at the center of the graph whereas the least connected ones are at the corners. Example :- Maine is connected to only one state and so it is in the corner.
• Connections between the states can be easily found out and number of hops between them can be easily calculated.
• All the states from the same regions are with the same color.

All the figures are done using excel.

##Rule of Thumb

No unjustified 3D: Occlusion hides information : Some of the objects cannot be seen becomes of other objects. The height or are of one attribute covers other attribute.

###Bad Design

To show a bad design I used 3D column plots. Here the count of all the categories of students in an institution is taken in different years. Each category of students is represented by differenct colored blocks and year on x-axis and count on y-axis.Z-axis has grad,undergrad, Phd,High School and Secondary school.

Dinesh Kumar Paladhi; CS 725:Information Visualization; Spring 2016; Computer Science; Old Dominion University

Data source used for all the graphs:-

http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_14_5YR_DP05&src=pt

Examples Referred :-

Mike Bostock (http://bl.ocks.org/mbostock/4063269) and Mike Bostock (https://bl.ocks.org/mbostock/3885211)