saifuddin778

We have a park with some shops in it. People visit our park and shop. We track visitors.

• We plot visitors as dots, and color code them by their ethnicity/race.
• We divide visitors in quadrants of gender and age (i.e. male-young, female-young, female-old, male-old).
• We compare visitors' family-size vs their number of purchases (in some lost hope) [sub plot # 1]
• We view visitors' purchasing history, looking for how much they have shopped in the last 5 months with us. [sub plot # 2]
• We classify visitors' purchases by product categories during last 5 months. [sub plot # 3]

Boring description aside, the main purpose of this plot is to demonstrate action binds (and resulting transforms) between different containers having independent svg spaces. Also uses Ziggy Jonsson's d3.legend for quick legend drawing.

saifuddin778

A Coxeter Plane in the form of layered squares. Each layer is "fully connected" with the preceding layer.

saifuddin778

A hypercube graph. Dedicated to a friend departed.

saifuddin778

A little manipulation while building the Greek Cross results in a magnificent Sierpiński triangle!

saifuddin778

A Greek Cross.

saifuddin778

Rhombic Chess + some interaction.

saifuddin778

Animated Quadtree.

A quadtree is a structure in which each node has less than or equal to 4 children.

From the visual 2-d perspective, it is usually illustrated in the form of quadrants, which follows the same strategy: if space gets more than 4 children, it is subsequently divided into 4 sub-quadrants or spaces.

saifuddin778

Also related to the Color Clustering - I, but this example (with a few parameters tweaking) achieves clustering with more well-defined and kind of hard-imposed hyperplanes between different color clusters, making the space more well distributed.

Observing visually, one sees how colors fight and take over each other, consume the space, and grow their colonies.

saifuddin778

This example achieves locality-sensitive color clustering by mimicing the behavior of Kohonen's Self Organizing Maps (a type of a neural network, commonly referred as SOM). With few modifications to Kohonen's idea, it is being demonstrated that how SOMs achieve a map of locality-sensitive soft data clustering, which is used in tasks like unsupervised (and supervised) modeling.

saifuddin778

This type of chained transition can be better understood as a ping-pong process: function A changes/moves some objects, and calls function B (A throws the ball at B). In return B changes/moves some objects, and calls function A (B throws the ball at A). This process keeps on going, until (of course) some interruption does happen.

This example transitions some random boxes in this fashion.