Lines Matching refs:collisions
122 # Various tests for hashing of tuples to check that we get few collisions.
125 # Earlier versions of the tuple hash algorithm had massive collisions
145 collisions = nballs - len(c)
146 z = (collisions - mean) / sdev
149 got = (collisions, pileup)
160 msg += f"coll {collisions:,} z {z:+.1f}"
204 # there's nothing the tuple hash can do to avoid collisions
205 # inherited from collisions in the tuple components' hashes.
215 # collisions among the hashes of the integers in L alone.
221 # Used to suffer a massive number of collisions.
234 # collisions is no more than 4 sdevs above the theoretically
452 # "pileup" is 0 if there were no collisions across those hash codes.
457 # "mean" is the number of collisions a perfectly random hash function
460 # "coll" is the number of collisions actually seen.
463 # of collisions a perfectly random hash function would suffer. A
495 # In a statistical sense that's waaaaay too many collisions, but (a) 324
496 # collisions out of a million hash codes isn't anywhere near being a