Refactoring of pyflagser functions into an unweighted and a weighted version.

[gtauzin]

Reference Issues/PRs

Fixes #27.

What does this implement/fix? Explain your changes.

Following on- and off-line discussions with @ulupo regarding the handling of zero values of the flag_matrix (#27), we have converged to the following solution implemented in this PR:

• the flagser function is to be split into two functions flagser_unweighted and flagser_weighted.
• flagser_unweighted focuses on the computation of homology. The input adjacency_matrix handles 0s as follows: Diagonal elements are vertex weights with non-0 or True values corresponding to True values and 0 or False values corresponding to False values. Off-diagonal, 0 or False values denote edge absence while non-0 or True values denote edges presence.
• flagser_weighted focuses on the computation of persistent homology. he input adjacency_matrix handles 0s as follows: If the matrix is a dense np.ndarray, zero values denote zero-weighted edges. If the matrix is a sparse scipy.sparse matrix, explicitly stored off-diagonal zeros and all diagonal zeros denote zero-weighted edges. Off-diagonal values that have not been explicitly stored are treated by scipy.sparse as zeros but will be understood as infinitely-valued edges, i.e., edges absent from the filtration.
• flagser_weighted now takes a max_edge_weight argument. All edge weights greater than that value will be considered as infinitely-valued, i.e., absent from the filtration. Additionally, it sets the maximum death values of diagram points. This parameter default is the maximum value allowed by the adjacency_matrix dtype.
• the IO function saveflag and load flags are each split into two functions: a unweighted and a weighted function than handle zeros in a way that is consistent with their flagser counterparts

Additionally,

• The two flagser functions code is now fully documented
• Handling of default parameters have been improved
• Sparse matrix efficiency warnings has been turned off (lil_matrix cannot be used because it ignores explicitely set 0 values).

[ulupo]

Thanks @gtauzin! Will review ASAP.

[gtauzin]

Putting it as WIP.

flagio also needs to be refactored accordingly.

[gtauzin]

I propose a consistent design for the static and persistence version of save and load functions too.

@ulupo Please let me know what you think.

[ulupo]

/azp run

[azure-pipelines[bot]]

Azure Pipelines successfully started running 1 pipeline(s).