-
Notifications
You must be signed in to change notification settings - Fork 125
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
TopoMap #987
TopoMap #987
Conversation
…algorithm. It computes a 2D embedding of points whose distance matrix is given as input. The embedding is computed so that the distances of a minimum spanning tree compute with Kruskal's algorithm are preserved. This commits includes a separate vtk module for this algorithm.
the TopoMap module to compute the embedding of the input points, given as coordinates or given their distance matrix.
… for the vtk layer, indicating for each point the time it was embedded in 2D.
…er can now choose between this one and Kruskal (the legacy one).
b657258
to
1358c43
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
hi @atalon-lip6,
thanks a lot for the PR.
here's a first batch of comments (more to come soon).
thanks for addressing them!
cheers,
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK, here is a second batch of remarks.
Thanks!
b642dc3
to
b9a4fb5
Compare
5a9b502
to
a2ee32b
Compare
…ted by the object).
a2ee32b
to
e62db22
Compare
alright, looks good. thanks! |
This PR introduces the Topological Mapper (TopoMap) method for DimensionReduction. TopoMap projects high-dimension points into 2D by computing and using a Minimum Spanning Tree (MST) on the high dimension points. We project the points so as to preserve the costs of the eges of the spanning tree.
It can use either Boost or Qhull for building the MST, but for now some bugs have been reported in peticular cases with Boost. A fix has been identified, yet not merged into Boost::Geometry for now.
This method was studied in the following paper: "Topomap: A 0-dimensional homology preserving projection of high-dimensional data" by Harish Doraiswamy, Julien Tierny, Paulo J. S. Silva, Luis Gustavo Nonato, and Claudio Silva, in
Proc. of IEEE VIS 2020. IEEE Transactions on Visualization and Computer Graphics 27(2): 561-571, 2020.