This is a fork of the mapper-tda library by Kiran Sanjeevan.
This fork allows the use of the DBSCAN clustering algorithm for mapper; I believe it is easier to heuristically determine the parameters for DBSCAN than it is for SLC, plus DBSCAN discards noise.
Additionally, it includes a filter function which treats the data as if it were a set of trajectories sample from some vector field, and then filters based on the distance along each trajectory. I think that this may be useful for applying topological NLDR methods to vector field data.