Hi,
I have a problem about the function BallsAndBins in main.cpp, m denotes the number of bin, n is the number of balls, s is the binScaler, it satisfies m=n*binScaler , but it seems that the code auto m = n / mm in line 373 is wrong. should it be auto m = n * mm?
what's more, Given the number of the bin b, the number of balls n, and one hash function,PSZ16 shows how to get the max binsize according to Equation(3) in Page13(simple hash).
But now if I use two hash function to map n balls into b bins(simple hash). each ball appears at two bins, now if I use Equation(3) to find the max binsize, should I plug the number of balls2*n and b in the Equation(3)? because the number of hash functions is 2, therefore the real number of balls turns out to be 2*n.
Generally,if I use k hash functions(simple hash), should I plug the ballsk*n and b in the Equation(3)? suppose b is fixed.
Hi, I have a problem about the function
BallsAndBins
inmain.cpp
,m
denotes the number of bin,n
is the number of balls,s
is the binScaler, it satisfiesm=n*binScaler
, but it seems that the codeauto m = n / mm
in line 373 is wrong. should it beauto m = n * mm
?what's more, Given the number of the bin b, the number of balls n, and one hash function,PSZ16 shows how to get the max binsize according to Equation(3) in Page13(simple hash).
But now if I use two hash function to map n balls into b bins(simple hash). each ball appears at two bins, now if I use Equation(3) to find the max binsize, should I plug the number of balls
2*n
andb
in the Equation(3)? because the number of hash functions is 2, therefore the real number of balls turns out to be2*n
.Generally,if I use k hash functions(simple hash), should I plug the balls
k*n
andb
in the Equation(3)? suppose b is fixed.