28 Dec
2017
28 Dec
'17
11:38 p.m.
On 12/22/2017 07:31 PM, Vitaly Magerya wrote:
On 12/22/2017 10:09 AM, Richard B. Kreckel wrote:
If the answer to the last two questions is yes, then we might add a stop condition and when that has been reached, try factorizing with an alternative permutation of symbols.
This may be a reasonable workaround (if it doesn't affect the non-hanging cases), but I'd personally be wary of treating a symptom without understanding the underlying cause.
Well, I'm unconvinced that this failure is some kind of bug in the implementation. If it's a problem with the algorithm, it could be worth trying alternative permutations of symbols. (And, of course, it shouldn't cause regressions.) -richy. -- Richard B. Kreckel <https://in.terlu.de/~kreckel/>