Dear Miguel,
Memory is not the sole limitation here. The main problem is that the worst
case time complexity is O(n^4). For instance, it took me several weeks of
computation to determine the hyperbolicity of the latest CAIDA maps (2012
and 2013).
There is currently no exact algorithm for comput
Vincent Delecroix wrote:
I was a bit wrong with my computations. Have a look at the file
SAGE_ROOT/src/sage/graphs/distance_all_pairs.pyx
You will see that matrices for distances are implemented as unsigned
short (whose maximum is 65535=2^16) while eccentricity is made of int
(whose maximum depe
