This library is my attempt at implementing an advanced Conway's Game of Life in Rust. My goal was to optimize and overengineer it in order to learn Rust along the way.

Conway's Game of Life (in my advanced version) plays in a grid containing cells (integers) that are either dead (0) or alive (>0). Then, the next generation is calculated with the prior time step following rule:

1. A cell is alive if its value is state.
2. Neighbors are either all eight surrounding cell (Moore) or only the four adjacent cells (VonNeumann).
3. If a cell was dead, it will be revived if it had a number of living neighbors specified in birth.
4. If a cell was alive, it will stay alive if it had a number of living neighbors specified in survive. If that's not the case, its value will decrease by one.

You can compile the program with cargo build --release. The binary then takes the following syntax:
gameoflife -{flags} {gif|tui}

• gif saves the Game of Life as a GIF with plotters,
• tui plots the Game of Life in the terminal with termion
• no command starts a dialogue that will ask about all parameters with sane defaults The flags can be:
• -a: algorithm (std or conv)
• -n: neighbor algorithm (m or vn)
• -i: number of iterations
• -x: number of columns
• -y: number of rows
• -p: probability of a cell being alive in the initial field
• -s: state
• -h: list all commands

The standard algorithms iterates over every cell, counts its neighbors, and then decides whether it's alive in the next step via normal if statements. This is fairly quick, especially as I used rayon to do this with multiple threads. The convolution algorithm, however, is about 3 faster. It uses ndarray-ndimage to convolve the field with the kernel [[1, 1, 1], [1, 0, 1], [1, 1, 1]] (Moore) or [[0, 1, 0], [1, 0, 1], [0, 1, 0]] (VonNeumann), which is somehow extremely fast with only one thread (props to Nil!) and then calculates the next field with functional-style maps and addition, multiplication, and comparisons.

For all licenses, look into license.html.
This file was automatically created using cargo-about (Embark Studios).