Desmos Projects

I adapted V3 to accept 2 inputs (x, y pixel coordinates) and output rgb values. I wrote a python program to convert image files into Desmos list data. 

This is my third iteration of my neural network. Instead of a fixed structure, it allows for variable structure by recursively feed forwarding through the network. I also implemented momentum and batch gradient descent, and plan to add RMSProp for dynamic learning rate.

This is my second version of my neural network in Desmos. It approximates a 1-variable function and has 3 layers with 2 neurons each. It also features a loss graph.

I used the 3D graphing calculator to visualize the loss landscape for a network with one weight and one bias (x and y axis), with loss on the z axis. 

This is a tremendously improved version of v2 in terms of performance. I use a recursive function for the DDA algorithm instead of computing all grid intersections within a range. This saves a lot of computation as you aren't finding intersections beyond walls.

This is a 2D grid-based raycast renderer that employs the DDA algorithm to find intersection points in a grid. This along with texture data is used to render a Wolfenstein-3D-esque scene in real time. Currently, I am working on sprite rendering.

This is a simple color-coded bubble sort algorithm implementation and visualization in Desmos. 

This was my first Desmos project. It's a simple bug animation that uses a custom procedural footplanting algorithm to move the legs realistically.

This is a snake game highly inspired by Google's snake game. The snake is a represented by list of points and uses the ticker to update inputs, position, and apples. Currently, there is no way to lose.