Q: Can RIVET handle arbitrary bifiltrations?
Yes. In particular, it now handles "multi-criticial" bifiltrations (i.e., ones in which a simplex is born at multiple incomparable grades)
Q: How large a bifiltration can RIVET handle?
: The computation time and memory footprint of RIVET are controlled by coarsening the bifiltration, so that the bigrades of appearance of simplices in the bifiltration live on a $k_x$ x $k_y$ grid for some small values of $k_x$ and $k_y$. In the initial release of RIVET, the runtime typically increased very quickly as $k_x$ and $k_y$ increased. In the current version of RIVET, this dependence is much milder, and one can get away with significantly larger values of $k_x$ and $k_y$.
Coarsening our bifiltrations at bit (say $k_x=k_y=50$), we have been able to comfortably handle bifiltrations coming from real data with more than 150 million simplices. For example, we recently used RIVET to analyze the first persistent homology of a full Vietoris-Rips bifiltration with 1000 points. On a computer with a lot of RAM, the current version RIVET may be able to handle significantly larger examples than that; we have not invested much time in exploring the limits.
Whatever those limits are, we expect RIVET to continue to become much faster and more memory efficient in the near future, as we continue to optimize the code.
Visualizing 0th persistent homology modules in RIVET is often much less expensive than for higher degree homology modules, and 0-th persistent homology is often more interesting in the two-parameter setting than in the ordinary one-parameter setting. So if your data set is large, you might try RIVET first with a 0th homology persistence computation.
Q: I'm interested in working on the development of RIVET or, more generally on the computational aspects of multi-D persistent homology. Is there room for me to get involved?
: Yes. There is a huge amount to do, both on the math side and the development side. We'd be happy to talk to you about this.