Class HartLA63


  • public class HartLA63
    extends FactorAlgorithm
    Experimental Hart algorithm assembling square congruences from smooth congruences. For small N, parameter optimization leads to C=0 and maxQRestExponent=0, indicating that the square test is the only thing that matters and such that the orginal Hart algorithm is much better. For larger N, CFrac seems to be certainly faster.
    Author:
    Tilman Neumann
    • Constructor Detail

      • HartLA63

        public HartLA63​(float C,
                        float maxQRestExponent,
                        TDiv_CF63 auxFactorizer,
                        int extraCongruences,
                        MatrixSolver matrixSolver)
        Standard constructor.
        Parameters:
        C - multiplier for prime base size
        maxQRestExponent -
        auxFactorizer - the algorithm to find smooth Q
        extraCongruences - the number of surplus congruences we collect to have a greater chance that the equation system solves.
        matrixSolver - matrix solver for the smooth congruence equation system