B = 40 -7 1 μ_{21} = 819/1650, μ_{31} = 382/1650, μ_{21} = 13842/45339, 19 -8 3 9 -3 1while ||b_{2}^{*}||^{2} = 45339/1650 and ||b_{1}^{*}||^{2} = 1650. The proof of Theorem 5.1 breaks down for these numbers, but nevertheless, b_{3} = [9, -3, 1] is a (unique) shortest multiplier.
B = -1 1 0 1 2 -2 -1 0 1Then b_{3} = [-1, 0, 1] and b_{3} - b_{1} = [0, -1, 1] are the shortest multipliers.
B = 3 -4 1 μ_{21} = -7/26, μ_{31} = 1/2, -10 -3 11 μ_{32} = -2899/5931. 6 0 -5Then b_{3} = [6, 0, -5], b_{3} - b_{1} = [3, 4, -6] and b_{3} + b_{2} = [-4, -3, 6] are the shortest multipliers.
if there is j such that a_{k,j} ≠ 0 then
col2 ← least j such that a_{k,j} ≠ 0;
else col2 ← n+1;
with
if there is j such that a_{k,j} ≠ 0 then
col2 ← least j such that a_{k,j} ≠ 0;
if a_{k,col2} < 0 then Minus(k); a_{k} ← -a_{k}; b_{k} ← -b_{k};
else col2 ← n+1;
The need for such a correction was pointed out by Benjamin Hilprecht.
Last modified 18th February 2017