Computing the Determinant of a Dense Matrix over  \({\mathbb Z}\)

We present a new, practical algorithm for computing the determinant of a non-singular dense, uniform matrix over \({\mathbb Z}\) ; the aim is to achieve better practical efficiency, which is always at least as good as currently known methods. The algorithm uses randomness internally, but the result is guaranteed correct. The main new idea is to use a modular HNF in cases where the Pauderis–Storjohann HCOL method performs poorly. The algorithm is implemented in OSCAR 1.0.