The game_of_life.py script in this directory calculates new generations of "dead" and "alive" cells according to the rules of the famous cellular automaton. Only a small part of the script is related to the actual rules for calculating a new generation:
kernel = np.array([
[1, 1, 1],
[1, 0, 1],
[1, 1, 1],
], dtype=np.uint8)
nr_alive_cells = lfr.focal_sum(generation, kernel)
underpopulated = nr_alive_cells < 2
overpopulated = nr_alive_cells > 3
reproducing = nr_alive_cells == 3
generation = lfr.where(generation, ~(underpopulated | overpopulated), reproducing)Such a script can be written and read by anybody with some experience with map algebra or array
algebra. Note that bitwise operators are being used for performing logical operations: ~ is
logical not, | is logical or (and & is logical and). In Python it is not possible to overload
not, or, and and.
Even though the script looks like any map algebra or array algebra script, the calculations are performed in parallel using a user-defined number of CPU cores (by default all of them). Also, computations continue at the same time as I/O is being performed.
In case you are interested in calculating generations for very, very large arrays then you can. LUE scripts run unchanged on a standard MPI cluster.
See also:
